cyberark

Week in review: Acrobat Reader flaw exploited, Claude Mythos offensive capabilities and limits Here’s an overview of some of last week’s most interesting news, articles, interviews and videos: Bringing governance and visibility to machine and AI identities In this Help Net Security interview, Archit Lohokare, CEO of AppViewX, explains how the rise of AI marked a turning point where machine and AI agent identities began converging into a single problem. Drawing on his experience across IBM and CyberArk, he describes the shift from human-driven systems to autonomous machines. MITRE releases a shared fraud-cyber framework built from real attack data The MITRE Fight Fraud Framework, known as F3, is a behavior-based model designed to give both teams a common structure for describing, detecting, and disrupting fraud campaigns. F3 organizes fraudster behavior into tactics and techniques drawn from real-world incidents. The tactics cover the full attack lifecycle: Reconnaissance, Resource Development, Initial Access, Defense Evasion, Positioning, Execution, and Monetization. ZeroID: Open-source identity platform for autonomous AI agents ZeroID is an open-source identity platform that implements an identity and credentialing layer specifically for autonomous agents and multi-agent systems. The core issue ZeroID targets is attribution in agentic workflows. When an orchestrator agent spawns sub-agents to carry out parts of a task, each sub-agent may call APIs, write files, or execute shell commands. Fixing vulnerability data quality requires fixing the architecture first In this Help Net Security interview, Art Manion, Deputy Director at Tharros, examines why vulnerability data across repositories stays inconsistent and hard to trust. The problem starts with systems not designed to collect or manage that data well. They introduce the idea of Minimum Viable Vulnerability Enumeration (MVVE), a minimum set of assertions needed to confirm two systems describe the same vulnerability, and find no true minimum exists. Review: The Psychology of Information Security Security controls fail when they are designed without regard for the people who must use them. That is the central argument of Leron Zinatullin’s second edition, and it is an argument he builds methodically across 17 chapters that draw from organizational psychology, change management, and usability research. Agentic AI memory attacks spread across sessions and users, and most organizations aren’t ready In this Help Net Security interview, Idan Habler, AI Security Researcher at Cisco, breaks down a threat most security teams haven’t named yet: agentic memory as an attack surface. Habler walks through MemoryTrap, a disclosed and remediated method to compromise Claude Code’s memory, showing how a single poisoned memory object can spread across sessions, users, and subagents. Network segmentation projects fail in predictable patterns Most enterprise networks have segmentation on the roadmap. Many have had it there for years. A survey of 400 U.S.-based network security practitioners who lived through failed segmentation projects finds that failure clusters into four distinct patterns, and the type of failure a team experiences depends heavily on the kind of environment and approach they attempted. Coordinated vulnerability disclosure is now an EU obligation, but cultural change takes time In this Help Net Security interview, Nuno Rodrigues Carvalho, Head of Sector for Incident and Vulnerability Services at ENISA, discusses the recent CVE funding scare and what it exposed about the fragility of global vulnerability disclosure infrastructure. He outlines how EU regulations, including the Cyber Resilience Act and NIS2, are creating stronger accountability for vendors and organizations. The exploit gap is closing, and your patch cycle wasn’t built for this The Cloud Security Alliance has published a briefing on what it calls a turning point in the threat landscape: the time between a vulnerability being discovered and a working exploit is shrinking fast. EU cybersecurity standards are at risk if supplier ban passes The European standards body ETSI sent a formal position paper to the European Commission, calling for changes to the proposed Cybersecurity Act 2 (CSA2), the EU’s planned revision to its existing cybersecurity certification framework. GitHub lays out copyright liability changes and upcoming DMCA review for developers A U.S. Supreme Court ruling issued in March has settled a question that has circulated among platform operators and developers for years: whether a service provider can be held liable for copyright infringement committed by its users without evidence of intent to contribute to that infringement. Cargo theft malware actor spent a month inside a decoy network before researchers pulled the plug Proofpoint researchers executed a malicious payload from a threat actor known to target trucking and logistics companies in late February 2026, doing so inside a decoy environment. The environment stayed compromised for more than 30 days, long enough for researchers to watch the actor work through their tools, scripts, and decisions beyond the initial break-in. Workplace stress in 2026 is still worse than before the pandemic Roughly 40% of employees worldwide said they experienced a lot of stress during the previous day, according to Gallup’s State of the Global Workplace 2026 report, a figure that has remained above pre-pandemic levels for several years. Daily anger stood at 22% globally, sadness at 23%, and loneliness at 22%. Together, these numbers point to a workforce that has not returned to the emotional baseline it held before 2020. Adobe issues emergency fix for Acrobat Reader flaw exploited in the wild (CVE-2026-34621) Adobe has pushed out an emergency security update for Adobe Acrobat Reader, patching a zero-day vulnerability (CVE-2026-34621) exploited in the wild since November 2025. CVE-2026-34621 is a critical prototype pollution vulnerability – a type of vulnerability that occurs in JavaScript and allows attackers to add or modify an application’s JavaScript objects and properties. Hackers hijacked CPUID downloads, served STX RAT to victims If you tried to download software from CPUID’s website late last week, you might have downloaded malware instead. CPUID (at cpuid[.]com) is a website that hosts free software utilities, primarily for Windows and Android. Booking.com data breach: Customer reservation data exposed “Unauthorized third parties may have been able to access certain booking information associated with your reservation,” email alerts sent out by Booking.com over the weekend warn. The online travel agency did not say which system(s) were accessed by the unauthorized third parties nor explained the scope of the incident. Testing reveals Claude Mythos’s offensive capabilities and limits Could Claude Mythos Preview, Anthropic’s latest large language model, be leveraged for fully automated cyber attacks? The UK government’s AI Security Institute (AISI) tested its capability to successfully engage in capture-the-flag (CTF) challenges and multi-step attack scenarios, and found that that while its cybersecurity capabilities exceed those of previously available models, it can’t reliably execute autonomous attacks on hardened networks. Fortinet fixes critical FortiSandbox vulnerabilities (CVE-2026-39813, CVE-2026-39808) Two vulnerabilities (CVE-2026-39813, CVE-2026-39808) in FortiSandbox could be leveraged by unauthenticated attackers to bypass authentication and execute unauthorized code or commands on vulnerable systems. Both vulnerabilities can be triggered with a specially crafted HTTP request, putting unpatched FortiSandbox deployments at risk. NIST admits defeat on NVD backlog, will enrich only highest-risk CVEs going forward NIST is overhauling how it manages the National Vulnerability Database (NVD) and switching to a risk-based model that prioritizes “enrichment” of only the most critical CVE-numbered security vulnerabilities. Researcher drops two more Microsoft Defender zero-days, all three now exploited in the wild The security researcher who earlier this month published a proof-of-concept (PoC) exploit for a zero-day privilege escalation vulnerability in Microsoft Defender is back with two more. The first, dubbed “RedSun,” is another privilege escalation flaw in the same platform. The second, “UnDefend,” allows a standard user to block Microsoft Defender from receiving signature updates or disable it entirely (if Microsoft pushes a major Defender update). 29 million leaked secrets in 2025: Why AI agents credentials are out of control GitGuardian’s State of Secrets Sprawl Report found 28,649,024 new secrets exposed in public GitHub commits across 2025, a 34% year-over-year increase and the largest annual jump in the report’s history. Product showcase: Stop secrets from leaking through AI coding tools with GitGuardian AI coding assistants are becoming part of everyday development, but they introduce new risks: secrets can be exposed before code reaches a repository or CI pipeline. Developers may paste API keys into prompts, or AI agents may access sensitive data through files and commands. Once inside the workflow, that data can be sent to model providers, logged, or cached. GitGuardian addresses this with ggshield AI hooks, which scan prompts and actions in real time to detect and block secrets before they are exposed. Why manual certificate management is running out of time In this video, John Murray, Senior Vice President of Sales at GlobalSign, explains what’s changing in the certificate industry and what companies need to do about it. Certificate validity periods are shrinking, which means companies will need to rotate certificates far more often than before. Zero trust at year two: What nobody planned for In this Help Net Security video, Jim Alkove, CEO of Oleria, walks through where zero trust programs typically stand one to two years in. Most organizations have made gains in endpoint security and network segmentation, but identity remains the stubborn problem. Identity sprawl, legacy system exceptions, and workforce friction each contribute to stalls that few programs anticipated. Webinar: The IT Leader’s Guide to AI Governance Generative AI is moving into everyday enterprise use, often outpacing governance. As adoption grows, organizations face challenges around security, privacy, and control. This discussion explores how enterprises manage AI governance in practice, focusing on real-world tradeoffs. Learn how guardrails, trusted content, and API-first platforms like headless CMS help bring AI under control while maintaining speed and visibility. Google makes it harder to exploit Pixel 10 modem firmware Google is working to improve the security of Pixel phones by focusing on the cellular baseband modem, a part of the device that handles communication with mobile networks and processes external data. $12 million frozen, 20,000 victims identified in crypto scam crackdown More than $12 million has been frozen, and over 20,000 victims have been identified in an international law enforcement operation targeting cryptocurrency and investment scammers. Basic-Fit hack compromises data of up to 1 million members Basic-Fit, a European gym chain, disclosed that hackers breached one of its internal systems, exposing members’ personal data in several countries. The company operates more than 2,150 clubs in 12 countries under two brands, with more than 5.8 million members. W3LL phishing service sold for $500 dismantled by the FBI The W3LL phishing kit, a cybercrime tool used to impersonate legitimate login pages and steal usernames and passwords, has been dismantled by the FBI and Indonesian law enforcement authorities. Officials estimate the operation was tied to more than $20 million in attempted fraud. Microsoft ends desktop detour for sensitivity labels in Office web apps Microsoft is rolling out an update to Office for the web that removes a long-standing limitation around document protection, adding new control to browser-based apps. OpenAI expands its cyber defense program with GPT-5.4-Cyber for vetted researchers Defending critical software has long depended on the ability to find and fix vulnerabilities faster than attackers can exploit them. OpenAI is expanding a program designed to give professional defenders prioritized access to AI tools built for that purpose. Alongside that expansion, OpenAI is releasing GPT-5.4-Cyber, a version of GPT-5.4 fine-tuned specifically for defensive cybersecurity work. Windows is getting stronger RDP file protections to fight phishing attacks Microsoft has introduced new Windows protections starting with the April 2026 security update to reduce phishing attacks that abuse Remote Desktop (.rdp) files. With these updates, the Remote Desktop Connection app displays stronger warning dialogs before a connection is established, shows details about the remote system, and requires users to review any request to share local resources such as drives or the clipboard. European AI spending set to hit $290 billion by 2029 European enterprises are committing serious money to AI, and the numbers are accelerating. According to IDC’s Worldwide AI and Generative AI Spending Guide, AI spending across Europe will reach $290 billion by 2029, growing at a compound annual growth rate of 33.7%. Command integrity breaks in the LLM routing layer Systems that rely on LLM agents often send requests through intermediary routing services before reaching a model. These routers connect to different providers through a single endpoint and manage how requests are handled. This layer can influence what gets executed and what data is exposed. A recent study examined 28 paid routers and 400 free routers used to access model APIs. Anthropic tests user trust with ID and selfie checks for Claude Anthropic announced identity verification for Claude using government ID and selfie checks, becoming the first major AI chatbot to do so, a move that may prove unpopular with users. Having built its reputation around privacy in the AI race, Anthropic risks undermining its positioning, as competitors such as OpenAI’s ChatGPT and Google’s Gemini do not require such verification. Two US nationals jailed over scheme that generated $5 million for the North Korean regime Two US nationals have been sentenced for their role in a scheme that placed North Korean IT workers inside American companies under false identities. Over several years, the operation used stolen identities from at least 80 US individuals and brought in more than $5 million for the North Korean government. Anthropic releases Claude Opus 4.7 with automated cybersecurity safeguards Software teams building agentic AI workflows have been pushing frontier models toward longer, unsupervised task runs. Claude Opus 4.7, now generally available from Anthropic, is aimed squarely at that demand, with particular gains in software engineering, multimodal processing, and the kind of instruction fidelity that matters when a model is running tasks autonomously over multiple steps. Social media bans might steer kids into riskier corners of the internet Governments are moving to block children under 16 from social media in the name of safety. But once these measures move from policy to practice, they raise a harder question: what happens when protecting kids requires collecting more data than ever before and may put them at greater risk? Apple AirTag tracking can be misled by replayed Bluetooth signals Apple’s AirTag is designed to help users track lost items by relying on a vast network of nearby Apple devices. New research shows that this same system can be manipulated to display locations where an AirTag has never been. Android 17 Beta 4 arrives with post-quantum cryptography and new memory limits Google shipped Android 17 Beta 4 on April 16, marking the last scheduled beta in the Android 17 release cycle. The build targets app compatibility testing and platform stability ahead of the final release, and it carries several behavior changes that developers need to account for before the stable version ships. Mozilla challenges enterprise AI providers with Thunderbolt, open-source AI client under your control For organizations that want to keep company data within their own systems and have more control over how AI is deployed, Mozilla is offering an alternative to externally hosted AI services with Thunderbolt, an open-source AI client designed for self-hosted use. Google wipes out 602 million scam ads with Gemini on duty Google claims that its security teams work around the clock using its Gemini AI models to detect and stop harmful ads. Malvertising remains an ongoing issue across Google’s ad network, with attackers abusing paid ads to pose as legitimate brands and lure users into malware downloads or phishing sites. The fully free Linux OS Trisquel gets a major update with version 12.0 Ecne Trisquel GNU/Linux, a free operating system aimed at home users, small enterprises, and educational centers, released version 12.0. The release, codenamed Ecne, is declared production-ready and builds on the previous version, Aramo, with changes to packaging, the kernel, security, and available software. Seized VerifTools servers expose 915,655 fake IDs, 8 arrested On April 7 and 8, Dutch police arrested eight suspects in a nationwide operation targeting users of the VerifTools platform as part of an identity fraud investigation. The suspects, all men aged 20 to 34, are accused of identity fraud, forgery, and cybercrime-related offenses. During searches, officers seized smartphones, laptops, cash, cryptocurrency, and weapons or items resembling them. AI adoption is outpacing the safeguards around it The 2026 AI Index from Stanford’s Institute for Human-Centered Artificial Intelligence outlines the broader environment around AI growth, including economic value, labor market effects, and the role of AI sovereignty. It also examines developments in science and medicine, the saturation of benchmarks, and governance frameworks that are struggling to keep up. Google to penalize sites that hijack the back button Google is broadening its spam policies to crack down on “back button hijacking,” a deceptive practice where websites interfere with browser navigation, blocking users from returning to the page they came from. DavMail 6.6.0 patches a regex flaw and advances its Microsoft Graph backend Organizations that run DavMail to bridge standard mail clients to Microsoft Exchange or Office 365 received an update this week. Version 6.6.0 addresses a code-scanning alert tied to a regex vulnerability, adjusts OAuth redirect handling to match a recent Microsoft change, and ships fixes across IMAP, SMTP, CalDAV, and CardDAV subsystems. OpenSSL 4.0.0 release cuts deprecated protocols and gains post-quantum support OpenSSL 4.0.0 removes several long-deprecated features, adds support for Encrypted Client Hello, and introduces API-level changes that will require code updates for applications built against older versions. Legitify: Open-source scanner for security misconfigurations on GitHub and GitLab Misconfigured source code management platforms remain a common entry point in software supply chain attacks, and organizations often lack visibility into which settings put them at risk. Legitify, an open-source tool from Legit Security, addresses that gap by scanning GitHub and GitLab environments and reporting policy violations across organizations, repositories, members, and CI/CD runner groups. What changed in nginx 1.30.0 and what it means for your upstream config nginx 1.30.0 brings together features accumulated across the 1.29.x mainline series. The release covers a broad range of changes, from protocol support additions to security-relevant fixes and new configuration options. Raspberry Pi OS 6.2 disables passwordless sudo by default Raspberry Pi OS 6.2, based on the Trixie version, introduces small changes, bug fixes, and disables passwordless sudo by default for new installations. Wi-Fi roaming security practices for access network providers and identity providers Public Wi-Fi roaming networks carry authentication credentials across multiple administrative boundaries, and the protocols governing that process vary widely in their security properties. The Wireless Broadband Alliance published a set of guidelines that specifies which authentication, encryption, and credential-handling practices operators should apply to networks running Passpoint and OpenRoaming. Product showcase: Ente Auth encrypts, backs up, and syncs 2FA Ente Auth is a free, open-source authenticator app designed to generate and store one-time passcodes for 2FA. It supports setup through QR codes and manual entry, allowing users to add accounts and begin generating codes. OpenAI updates Agents SDK, adds sandbox for safer code execution OpenAI’s updated Agents SDK helps developers build agents that inspect files, run commands, edit code, and handle tasks within controlled sandbox environments. The update provides standardized infrastructure for OpenAI models, a model-native harness that lets agents work with files and tools on a computer, and native sandbox execution for running tasks safely. Google Play is changing how Android apps access your contacts and location Google’s new set of Google Play policy updates and account transfer feature strengthen user privacy and protect businesses from fraud. Google is also expanding features for managing new contact and location policy changes to support a smoother, more predictable app review experience. Codex can now operate between apps. Where are the boundaries? OpenAI is rolling out a major update to the Codex desktop app for users signed in with ChatGPT. Personalization features, including context-aware suggestions and memory, will roll out to Enterprise, Edu, and users in the EU and UK soon. Computer use is initially available on macOS and will expand to EU and UK users in the near future. Cybersecurity jobs available right now: April 14, 2026 We’ve scoured the market to bring you a selection of roles that span various skill levels within the cybersecurity field. Check out this weekly selection of cybersecurity jobs available right now. New infosec products of the week: April 17, 2026 Here’s a look at the most interesting products from the past week, featuring releases from Axonius, Broadcom, Siemens, and Sitehop.

Capsule Security debuts with $7 million funding to secure AI agent behavior Capsule Security has launched from stealth with a $7 million seed round led by Lama Partners and Forgepoint Capital International. It prevents AI agents from being manipulated, misbehaving, or silently exfiltrating data when handling sensitive information and executing workflows. Capsule is advised by security juggernauts who have put their trust in the company’s innovative approach, including Chris Krebs, the first Director of CISA; Omer Grossman, former Global CIO at CyberArk; Jim Routh, former CISO across multiple global, Fortune 500 enterprises; and Dr. Yonesy Núñez, former CISO and senior security executive across financial services. “AI agents are a new class of privileged user, operating at machine speed with minimal oversight,” said Chris Krebs. “Legacy tools weren’t built to monitor what happens between prompt and action—that’s the runtime gap. Capsule closes it.” AI agents are being introduced into enterprise environments at record speed, with permissions and direct paths into critical business systems. Microsoft recently reported that more than 80% of Fortune 500 companies now use active AI agents built with low-code and no-code tools. Concurrently, coding agents are swiftly expanding the speed and scope of automation. The recent rollout of Claude Code Security, for example, is being viewed as a wake-up call that the future of cybersecurity will be shaped by deep runtime and AI capabilities, not by posture dashboards alone. “AI agents are quickly becoming a new class of privileged user in the enterprise, except they can act at machine speed and they do not behave like deterministic software,” said Naor Paz, CEO of Capsule Security. “That creates a dangerous gap between what security teams can govern today and what agents can do in production. Capsule closes that gap by enforcing trust at runtime, inside the execution path, so teams can move fast with agents while staying in control of what those agents can access and execute.” ShareLeak, PipeLeak, and ClawGuard: proof points and real-world disclosures The risk is not theoretical. In open frameworks like OpenClaw, every tool invocation becomes a decision point and potential strike. To address this, Capsule created ClawGuard, an open-source enforcer that adds a pre-invocation checkpoint before agents execute tool calls. The same class of risk is emerging in mainstream agent platforms. Concurrent to its emergence from stealth, Capsule has published two research reports detailing zero-day vulnerabilities discovered in leading platforms, ShareLeak and PipeLeak. ShareLeak is a critical severity indirect prompt injection vulnerability in Microsoft Copilot Studio, patched and assigned CVE-2026-21520. PipeLeak, another prompt injection vulnerability, was discovered in Salesforce Agentforce, triggering through untrusted lead-form inputs to influence agent behavior and drive unsafe downstream actions. Together, the findings show how suspicious content can hijack agent goals and steer tool usage, turning routine workflows into high-impact risk paths. (Salesforce confirmed it has addressed the issue.) “The agentic AI boom is creating an opening in runtime behavior enterprises can’t afford to ignore,” said Capsule Advisor, Omer Grossman. “The ability to secure this layer is what ultimately determines whether companies can move fast with AI without breaking trust. That is why I chose to support Capsule Security. The team is addressing the problem at its core by delivering real-time visibility and control over agent behavior, grounded in the operational reality of AI-driven environments within a fundamentally new and rapidly evolving paradigm.” Capsule’s runtime models evaluate actions in context and can block unsafe or unauthorized activity before it completes. Capsule also generates auditable telemetry designed for governance, investigation and compliance teams. The platform is designed to secure both third-party and custom agent deployments without added infrastructure. No proxies, gateways, SDKs, or browser extensions required. Capsule supports Cursor, Claude Code, Microsoft Copilot Studio, ServiceNow, and Salesforce Agentforce and integrates with existing security workflows, enabling teams to route agent telemetry into established response processes. “Agents have the ‘superpower’ to write and deploy code at unprecedented rates, fundamentally changing how software is built and operated,” said Ron Zalkind, Founding General Partner at Lama Partners and Board Member at Capsule Security. “With that level of power comes a new responsibility to secure it. Security leaders understand that legacy tools were never designed to interpret intent, context, and real-time behavior, which are essential for securing dynamic agentic environments. From day one, Naor and Lidan have combined deep technical rigor with clarity of vision to build a platform that allows organizations to confidently adopt AI agents while stopping dangerous actions before damage is done,” Zalkind added. “Capsule fine-tuned Small Language Models (SLMs) to create a multi-agent system of ‘Guardian Agents’ that can protect AI with AI, covering both posture and low-latency runtime protection. The team is the strongest of the agent-space players, having expertise in both traditional security and deep familiarity with emerging protocols like MCP and Skills,” Damien Henault, Managing Director/Partner at Forgepoint Capital International and Capsule Board Member, concluded.

Bringing governance and visibility to machine and AI identities In this Help Net Security interview, Archit Lohokare, CEO of AppViewX, explains how the rise of AI marked a turning point where machine and AI agent identities began converging into a single problem. Drawing on his experience across IBM and CyberArk, he describes the shift from human-driven systems to autonomous machines. Lohokare also shares how AppViewX, together with Eos, is building a unified approach that combines CLM, PKI, and agentic governance to secure these identities. You’ve gone from building CyberArk’s next-generation Identity Security Platform to founding Eos to now leading AppViewX. At what point did you recognize that machine identity and AI agent identity were converging into a single, unsolvable problem without a unified platform? The realization came with the inflection point of AI adoption, specifically when tools like ChatGPT moved from novelty to enterprise reality. For the past two decades, enterprise digital transformation has been driven by humans using applications and infrastructure at scale. My work during that time, at IBM, Centrify, Idaptive and then CyberArk and beyond, was focused on solving the identity security challenges that emerged from that human-centric growth. But with the rapid rise of AI, we’ve entered a fundamentally different era. The next step-function in productivity isn’t coming from more humans, it’s coming from machines and AI agents operating autonomously and at massive scale. That shift made something very clear to me: we’re no longer dealing with separate problems of machine identity and emerging AI identity. They are converging into a single, exponentially growing challenge, one that cannot be solved with platforms designed for human identity. That was the moment of realization. The next generation of market-leading identity security platforms will be those purpose-built to secure machine and AI agent identities, not humans. AppViewX already had traction in CLM and PKI. Eos brings agentic governance on top of that. Walk us through what the integration of these two layers looks like in practice for a security engineer sitting at a SOC console. AppViewX already delivers deep discovery, governance, and compliance for machine identities through CLM, along with strong authentication via PKI. Eos extends this with an agentic layer, bringing the same capabilities to AI agents. The challenge doesn’t change, but rather, it scales. Organizations now need to discover, govern, and secure not just certificates, but autonomous agents accessing resources. For a security engineer, this becomes a unified control plane to: Discover machine identities and AI agents Enforce consistent governance and policy Monitor, audit and control their access and behavior The value is simplicity and consolidation: one platform to reduce identity and access management risk across both machines and AI agents as they rapidly scale. You and Kashyap Ivaturi are co-founders stepping into the top two technology roles of an established company. How are you thinking about preserving the engineering culture that made Eos attractive in the first place while integrating into a larger organization? At Eos, Kashyap was instrumental in building an AI-native development lifecycle (AI-DLC), leveraging tools like Claude Code, Figma Make, and other AI agents to dramatically accelerate innovation. What’s unique is that he’s done this across both small, high-velocity teams and large organizations of 500+ engineers, so he brings proven playbooks for scaling without losing that core culture. Importantly, AppViewX was already on a similar path. Even before we stepped in, there were strong engineering pods experimenting with and adopting AI-driven development practices. So culturally, there’s a natural alignment rather than a forced integration. Our focus isn’t to impose a new culture, but to amplify what’s already working. That means keeping teams small and empowered, doubling down on AI-native workflows, and creating shared platforms and guardrails that allow innovation to scale without adding unnecessary process. Ultimately, the goal is to preserve the speed and creativity of Eos, while combining it with the scale and discipline of AppViewX, so we get the best of both worlds. AI agents can spawn other agents, delegate privileges, and operate across trust boundaries in ways traditional workloads never did. Where do you see the most dangerous blindspot in how enterprises are currently governing this? The biggest blind spot is that enterprises are still applying human and workload-centric identity models to something fundamentally different. AI agents are dynamic, non-deterministic, and composite, they can spawn other agents, delegate privileges, and operate across trust boundaries. Yet most organizations still treat them like service accounts or API keys, with coarse, long-lived permissions and little visibility. That creates a dangerous gap in identity, context, and accountability. Enterprises often don’t know which agent is acting, what it has access to, or how that access propagates. The core issue is the lack of a unified control plane to govern identity, posture lifecycle, privilege, and behavior in real time. Until that exists, organizations will carry significant unseen risk as agentic systems scale. Both IDC and KuppingerCole have recognized AppViewX in their leadership reports. Yet the machine identity management category is getting crowded fast, with CrowdStrike, Venafi, and others moving aggressively. What does AppViewX do that those players structurally cannot? AppViewX is built as a purpose-driven, next-generation SaaS platform for machine identity lifecycle management, not an extension of an endpoint product or a legacy system. Many newer entrants approach this space adjacently, treating machine identity as a feature to enhance visibility or detection. That limits how deeply they can manage the full lifecycle. AppViewX is built on core primitives: discovery, issuance, governance, control, and compliance. The platform is designed to operate seamlessly across heterogeneous, vendor-neutral environments. With the addition of AI agent identity through Eos, we now unify CLM, PKI, and agent governance under a single control plane. That’s not something you can easily retrofit into an endpoint or detection product, it requires a fundamentally different architecture, which is where we have a long-term advantage.

Cybersecurity jobs available right now: March 24, 2026 Application Security Analyst Alignerr | USA | Remote – No longer accepting applications As an Application Security Analyst, you will review and analyze application security scenarios across code, APIs, and system behavior. You will classify vulnerabilities such as authentication flaws, injection risks, and business logic issues, and evaluate secure coding practices and remediation strategies. You will also help create and validate security-focused reasoning datasets that train AI to accurately assess application risks. Application Security Engineer E.ON Digital Technology | Germany | Hybrid – No longer accepting applications As an Application Security Engineer, you will support the development and rollout of the application security roadmap. You will work closely with development and operations teams to enforce security practices and provide guidance early in the software development lifecycle to identify and mitigate risks. You will collaborate with Technology Platform teams to integrate security tools and processes into CI/CD pipelines, and guide developers on secure coding practices while helping remediate vulnerabilities. Cloud Security Lead Tata Consultancy Services | Ireland | Hybrid – No longer accepting applications As a Cloud Security Lead, you will own and govern cloud security controls across public and hybrid environments, including identity and access management, network security, workload and data protection, configuration baselines, and logging. You will provide cloud security architecture assurance for new and evolving solutions, ensuring alignment with security standards, risk appetite, shared responsibility models, and regulatory requirements. Get weekly updates on new cybersecurity job openings. Subscribe here! Corporate Cyber Security Specialist Paysafe | Ireland | Hybrid – No longer accepting applications As a Corporate Cyber Security Specialist, you will As a Corporate Cyber Security Specialist, you will lead small to medium security projects and support major initiatives, including the expansion of Microsoft 365 security services. You will design, implement, and optimize security automation, including ASR rules and other M365 controls. You will manage and enhance core security tools such as CASB solutions and CyberArk PAM, and build and strengthen IAM/IGA capabilities (Entra ID, SailPoint, Saviynt, Pathlock), including JML automation and access governance. Cyber Security Analyst The Cambridge Building Society | United Kingdom | Hybrid – No longer accepting applications As a Cyber Security Analyst, you will monitor and analyze outputs from security tools to identify early indicators of compromise. You will respond to cyber events and incidents, recommending and taking action to protect systems and data. You will fine-tune and enhance existing security controls to maximize their effectiveness, and coordinate regular testing, audits, and external security assessments, ensuring findings are risk-assessed and remediated within agreed timescales. Cybersecurity Automation Engineer Assurant | India | On-site – No longer accepting applications As a Cybersecurity Automation Engineer, you will design, implement, and maintain security solutions aligned with Assurant’s Information Security Policies and Standards. You will support technologies across applications, infrastructure, and endpoint security, and identify, assess, and remediate hardware and software vulnerabilities. You will ensure the confidentiality, integrity, and availability of enterprise systems, develop and recommend practices to prevent unauthorized access or data loss, and measure, analyze, and report compliance with security policies to leadership. Cybersecurity SOC L3 ZENDATA Cybersecurity | UAE | On-site – No longer accepting applications As a Cybersecurity SOC L3, you will lead SOC operations with a focus on operational excellence, visibility, and responsiveness. You will manage and mentor SOC analysts, fostering skill development and collaboration. You will drive SOC optimization through process improvements, automation, and technology adoption, and build and execute a strategic roadmap aligned with evolving threats and business goals. Head of Cyber Security Aneurin Bevan University Health Board | United Kingdom | On-site – No longer accepting applications As a Head of Cyber Security, you will provide strategic leadership and operational oversight for the organization’s cyber security posture, ensuring the protection of digital assets, clinical systems, and sensitive information. You will be responsible for developing and implementing cyber security strategy, managing risk, leading incident response, and ensuring compliance with national standards, regulatory requirements, and NHS Wales policies. Engineering Manager, Software Supply Chain Security: Pipeline Security GitLab | Canada | Remote – No longer accepting applications As an Engineering Manager, Software Supply Chain Security: Pipeline Security, you will guide the design and delivery of software supply chain security features, with a focus on CI job artifact security. You will lead the implementation of the SLSA framework in GitLab CI/CD and integrate capabilities such as SBOM, software composition analysis, and vulnerability management. Engineer – Network and Security KATIM | UAE | On-site – No longer accepting applications As an Engineer – Network and Security, you will architect, design, and implement scalable, resilient network solutions. You will deploy and manage network devices, including routers, switches, firewalls, and load balancers, ensuring availability, reliability, and performance meet business requirements. You will develop and enforce security policies, protocols, and procedures to protect the organization’s network and data, and conduct regular security assessments and audits to address vulnerabilities and ensure compliance with industry standards. Global Head IT Security & Compliance Lonza | France | Hybrid – No longer accepting applications As a Global Head IT Security & Compliance, you will develop, implement, and oversee a strategic enterprise information security and IT risk management program, including establishing and enforcing security policies and standards across the organization. You will create and manage security and risk awareness training for employees and system users, facilitate IT risk assessments in collaboration with stakeholders, and ensure compliance with relevant laws, regulations, and policies to minimize risk and audit findings. IT Security SecOps LSEG | France | Hybrid – No longer accepting applications As an IT Security SecOps, you will monitor cybersecurity events from internal tools and outsourced detection services. You will coordinate response and recovery efforts with IT teams, deliver cybersecurity service activities including SLA and KPI monitoring for insourced operations, and oversee outsourced security services to ensure effective performance and alignment. Penetration Tester Hacktive Security | Italy | Remote – No longer accepting applications As a Penetration Tester, you will lead and conduct comprehensive tests across systems, networks, web and mobile applications, and other digital assets. You will identify, assess, and ethically exploit vulnerabilities within client environments, and provide expert guidance on security best practices and risk mitigation to both technical and non-technical stakeholders. Safeguards Analyst, Account Abuse Anthropic | USA | Hybrid – No longer accepting applications As a Safeguards Analyst, Account Abuse, you will As a Safeguards Analyst, Account Abuse, you will develop and refine account signals and prevention frameworks that turn internal and external data into actionable abuse indicators. You will build and optimize identity and account-linking signals using graph-based infrastructure to detect coordinated, large-scale abuse. Security Engineer Paragon | Israel | On-site – No longer accepting applications As a Security Engineer, you will implement, configure, and maintain information security systems. You will continuously review and improve the organization’s infrastructure security architecture, ensuring that all requirements defined by the CISO and security team are properly implemented and aligned with the overall design. You will also actively monitor and analyze systems, firewalls, and logs to detect and respond to potential threats. Security Engineer Air Apps | USA | On-site – No longer accepting applications As a Security Engineer, you will develop and implement threat modeling to identify risks across applications and infrastructure. You will conduct vulnerability scanning, penetration testing, and security assessments to uncover weaknesses. You will define and enforce secure coding practices in collaboration with development teams and work with DevOps to integrate security into CI/CD pipelines and automate testing. Senior Cloud Security Consultant CyberCX | Australia | Hybrid – No longer accepting applications As a Senior Cloud Security Consultant, you will work with customers to design and deliver cloud security solutions, including native controls and third-party CSPM, CNAPP, or CIEM tools. You will assess the risk of existing cloud environments, workloads, and services against industry standards and frameworks, and provide recommendations to improve security posture. You will also prepare and present high-quality deliverables such as presentations, pre-sales proposals, assessment findings, and strategic recommendations. Senior Cloud Security Engineer Haventree Bank | Canada | Hybrid – No longer accepting applications As a Senior Cloud Security Engineer, you will act as the technical owner for key cloud security platforms, influencing configuration, detection logic, and roadmap in partnership with operations teams. You will define and maintain cloud security reference architectures across a multi-cloud environment, covering identity, network segmentation, encryption, workload protection, logging and monitoring, and secure service integration. You will also establish secure patterns for Infrastructure as Code, including secure-by-default templates, scanning expectations, and drift management. Senior Cyber Incident Responder Labcorp | USA | Hybrid – No longer accepting applications As a Senior Cyber Incident Responder, you will you will serve as the lead responder for validated cyber incidents, prioritizing threats that could impact clinical operations, EHR systems, connected medical devices, or PHI. You will coordinate with technical and clinical stakeholders to contain and remediate threats across hospitals, clinics, and remote care environments. You will drive improvements to the Incident Response Plan, ensuring readiness for ransomware, business email compromise, and other threats. Senior SAP ERP Security Specialist Merck Group | Germany | On-site – No longer accepting applications As a Senior SAP ERP Security Specialist, you will lead the security architecture across SAP S/4HANA and SAP ECC environments, ensuring alignment with enterprise security frameworks. You will manage user access governance and compliance, and integrate robust security controls into critical business processes. You will also monitor SAP systems for vulnerabilities, conduct risk assessments, and drive continuous improvement in security practices. SOC Analyst Tier 2 Elad Software Systems | Israel | Hybrid – No longer accepting applications As a SOC Analyst Tier 2, you will investigate and respond to escalations from Tier 1 analysts. You will conduct in-depth analysis of complex security alerts, including incidents, breaches, malware, phishing, and unauthorized access attempts. You will perform proactive threat hunting using advanced analytics, event correlation, and MITRE ATT&CK methodologies, and carry out endpoint forensics and network traffic analysis to identify attack vectors and scope. Vulnerability Analyst NTT DATA | Italy | On-site – No longer accepting applications As a Vulnerability Analyst, you will conduct vulnerability assessments using tools such as Nessus, Qualys, Rapid7, and OpenVAS. You will validate automated scan results to eliminate false positives, analyze vulnerabilities to assess exploitability and business impact, and create detailed reports with actionable remediation guidance. You will evaluate vulnerabilities using CVSS scoring and business context, prioritize remediation based on risk and asset criticality, and collaborate with threat intelligence teams to incorporate emerging threat data. Vulnerability Researcher Delta Dental Ins. | USA | Remote – No longer accepting applications As a Vulnerability Researcher, you will conduct research to identify high-impact, previously unknown vulnerabilities across a wide range of applications and technologies, including AI-enabled systems. You will perform vulnerability assessments using industry best practices across web applications, APIs, and cloud environments.

Passwordless authentication is often presented as the end of account takeover. But to understand the real threat landscape, we need to examine how passwordless is actually deployed in the real world. Attackers do not break protocols in theory. They target the most common implementations, the places where usability, scale and architecture intersect. Focusing on one of those common implementations, we examine Google Cloud Authenticator. This discussion explores the hidden mechanisms behind synced passkeys and their implementation within the Google ecosystem. Our aim is to help defenders better understand the technology, to lay the groundwork to show how new attack vectors could emerge in a passwordless environment. This post is Part 2 in our series examining passkey adoption from a security perspective. If you haven’t read Part 1 yet, we recommend starting here: The Art of the Invisible Key – Passkey Global Breakthrough. Palo Alto Networks customers are better protected from threats that take advantage of issues with cloud authentication through the following products and services: If you think you might have been compromised or have an urgent matter, contact the Unit 42 Incident Response team. When we set out to evaluate the security of passkeys, we deliberately thought like attackers. Instead of asking whether Fast IDentity Online (FIDO) is secure, we asked where passkeys live, how they move, how they sync and which components handle the most sensitive operations. That shift in perspective revealed a surprisingly broad and largely unexplored attack surface. Many of the findings we uncovered have not been publicly discussed, and we will reveal them throughout this series. However, before diving into new attack vectors, we need to establish a clear architectural foundation. The FIDO and W3C specifications define the authentication protocols in detail, but the real protection of key material often extends beyond those documents. In practice, critical implementation details are embedded in browsers, operating systems and cloud services, and are rarely described publicly. We therefore began with one of the most widely adopted passwordless ecosystems: Google’s passkey authentication. In this article, we examine the architecture behind synced passkeys for desktop users and explore the lesser known Google Cloud Authenticator, a cloud-based component that performs sensitive cryptographic operations. Once we understand how this system is built, we can analyze the new attack vectors it introduces and discuss how to mitigate them in the next part of this series. Disclaimer: This analysis reflects our understanding of a complex, evolving system, based on client code, runtime behavior, network traces and public sources. The research detailed here was conducted for responsible, ethical security analysis. To keep the discussion readable, we simplify certain internal flows and use illustrative pseudocode. Although the Google Cloud Authenticator is used by Chrome across platforms, our focus here is Chrome on Windows with Trusted Platform Module (TPM) support. Whenever users authenticate with passkeys backed by Google Password Manager (GPM) across desktop platforms (macOS, Windows, Linux and ChromeOS), we see a connection to the domain enclave.ua5v[.]com. As of January 2026, searching for enclave.ua5v[.]com yields surprisingly little public information about its role in passkey authentication (as shown in Figure 1). This is despite powering logins worldwide. The FIDO specifications do not explicitly define a cloud-based authenticator. However, related building-block elements exist in Client-to-Authenticator Protocol (CTAP) Hybrid transports, where Bluetooth Low Energy (BLE) physical proximity can be used to establish a tunnel service to Google’s caBLE.ua5v[.]com domain. While Chrome still leverages portions of the Hybrid (caBLE) transport code, understanding the actual implementation requires examining Chrome’s behavior and the cloud authenticator, as observed through its network interface and Chromium source code (as shown in Figure 2). Onboarding Device A Chrome user can perform passkey operations synchronized with their Google account, making passkeys available across all connected devices. Before any passkey can be used, Chrome runs a dedicated onboarding flow behind the scenes (shown in Figure 3). This allows the remote Google Cloud Authenticator to verify both the device’s identity and the user’s possession of it. To establish trust between the device and the cloud authenticator, Chrome assigns two TPM-backed key pairs: Identity key: Represents “something you have.” In WebAuthn terms: “Register a particular client device as a ‘trusted device’, so the client device itself acts as a something-you-have authentication factor for future authentication.” User verification key (UV key): Represents “something you know or are.” This key can only be created or used after the user authenticates (verifies) with the same method they use to unlock the device (biometric or PIN). After generating the device keys, the client sends a registration request to the cloud authenticator. The message includes: Commands: "device/register", "keys/genpair" Identity_public_key: Public key corresponding to the TPM-protected identity key. UV_public_key: Public key corresponding to the TPM-protected UV key. Device_id: SHA256 hash of the identity public key (SPKI). The cloud authenticator creates a new record and stores the device’s hardware-backed public keys associated with the device ID: devices[device_id] = { hw: identity_public_key, uv: uv_public_key } In addition, the cloud authenticator generates and stores a device-specific wrapping key. This key is used to encrypt secrets, allowing them to be stored on the device as opaque blobs and unwrapped only by the cloud authenticator: wrapping_keys[device_id] = random(32) Finally, the cloud authenticator generates a member key pair. The private member key is encrypted with the wrapping key. This key is then returned along with the public member key intended for joining the device as a trusted member within the account’s security domain of authorized devices: (member_private_key, a member_public_key) = Generate P-256 key pair wrapped_member_private_key = encrypt(member_private_key, key:wrapping_key) On the first device, the onboarding process also includes generating the account secrets: Security domain secret (SDS): A symmetric master key used by the cloud authenticator to encrypt and decrypt all synced passkeys for the account GPM PIN Code: A user-chosen secret that allows newly added devices to access the account’s synced passkeys Figure 4 shows the start of the recovery PIN process. Establishing a security domain backed by Google’s Trusted Vault service, linking the user’s authorized devices and managing the encryption keys used by Chrome Sync to securely synchronize passkeys Creating a PIN-protected recovery mechanism to store and recover the SDS securely During the first passkey operation on a new device or on a recovered account, Chrome prompts the user to verify with the same GPM PIN. The PIN is verified by the cloud authenticator and protected recovery mechanism. This allows the device to join the account’s security domain, synchronize account passkeys and enable the cloud authenticator to wrap the SDS for that device. To summarize the device onboarding process, we can review the various key materials generated during the onboarding and stored in a file under the user’s profile directory: %LocalAppData%\Google\Chrome\User Data\ \passkey_enclave_state. The local file enables future device-cloud communication without re-registration or re-entering the PIN and includes the following elements: Device keys: - Identity key: - wrapped_identity_private_key: When Chrome creates the identity key, it asks the TPM to seal the private portion with the TPM’s hard-coded key. This allows the key to be saved as an opaque blob that only that specific TPM can unseal. - identity_public_key: The corresponding public portion. - device_id: The hash of the identity public key that is used as a unique identifier for the device within the cloud authenticator - UV key - wrapped_uv_private_key: The label of the hardware-backed key that is gated by local (Windows Hello) user verification - uv_public_key: The corresponding public portion Identity key: wrapped_secret: The SDS encrypted with the cloud authenticator’s wrapping key wrapped_pin: PIN data encrypted under the cloud authenticator’s wrapping key, enabling the authenticator to verify the PIN, enforce retry limits and perform secure PIN updates without ever exposing the plaintext PIN Figure 5 shows the extracted passkey_enclave_state file. After the device onboarding — which involves completing enrollment with the cloud and joining the security domain — the device's user can start creating and using passkeys that are securely synchronized with their Google account. Figure 6 shows the flow for this process. When a user chooses to add a passkey as an authentication method to a service, the relying party (service) invokes a create WebAuthn API: navigator.credentials.create(options) Chrome then displays a prompt offering to save the passkey in GPM as shown in Figure 7. Once the user selects the GPM option, Chrome prepares the required data and initiates a secure, peer-to-peer (P2P) encrypted session with the cloud authenticator. Create Command (Chrome to Cloud Authenticator) Chrome sends a request containing the following parameters: command: "passkeys/create" device_id wrapped_secret Key Generation and Encryption (Cloud Authenticator to Chrome) The cloud authenticator performs the following operations: Uses the provided device_id to locate the corresponding stored wrapping_key. Unwraps the wrapped_secret to recover the SDS. Generates a new P-256 ECDSA key pair for the passkey. Encrypts the passkey’s private key using the SDS. Returns the public key and the encrypted private key to the Chrome client. wrapping_key = wrapping_keys[device_id] security_domain_secret = decrypt(wrapped_secret, key: wrapping_key) (passkey_private_key, passkey_public_key) = Generate P-256 key pair encrypted_private_key = encrypt(passkey_private_key, key:security_domain_secret) Return to the device (passkey_public_key, encrypted_private_key) Passkey Registration (Chrome to Relying Party) Chrome forwards the passkey public key to the relying party as part of the WebAuthn registration response. The website stores this public key under the user’s account for future authentication. Passkey Synchronization (Chrome to Security Domain) Next, Chrome prepares a protobuf-encoded sync entity named WebauthnCredentialSpecifics. This record represents the cloud authenticator’s encrypted view of the new credential, enabling any device enrolled in the same security domain to access and use it for authentication. Each WebauthnCredentialSpecifics entry includes: RP ID (relying party’s domain) Username Passkey public key Passkey encrypted private key Chrome uploads this sync entity to the Security Domain service, which distributes the update to the other registered devices. Whenever a new passkey is added to the account, each enrolled device stores the corresponding WebauthnCredentialSpecifics locally in Chrome’s sync database: %LocalAppData%\Google\Chrome\User Data\ \Sync Data\LevelDB. The stored record allows GPM to list the accounts' passkeys and make them available for authentication. Figure 8 shows the authentication flow. Log in With Synced Passkey Once a passkey has been created and synchronized, a user can initiate login to a relying party using the synced passkey from any enrolled device. The relying party then invokes a get WebAuthn API: navigator.credentials.get(options). Chrome locates the WebauthnCredentialSpecifics entity that matches the visited relying party ID and establishes a secure connection to the cloud authenticator. Assertion Request (Chrome to Cloud Authenticator) Chrome sends a request containing: Command: "passkeys/assert" client_data_json (challenge and rpID from WebAuthn request) device_id wrapped_secret WebauthnCredentialSpecifics The cloud authenticator performs the following operations: Uses the provided device_id to locate the corresponding wrapping_key Unwraps the wrapped_secret to recover the SDS Decrypts the passkey’s encrypted_private_key with the SDS Sets the authenticator flags, including the user-verified flag, based on whether the client’s message is signed with the user verification key (see the secure communication protocol in the Secure Communication Protocol section) Constructs authenticator_data, which includes: relying party ID (hash), flags and signature counter (always zero) Using the passkey_private_key, signs the concatenation of the client_data_json and the authenticator_data Returns to the client AuthenticatorAssertionResponse containing the client_data_json, authenticator_data and the signature wrapping_key = wrapping_keys[device_id] security_domain_secret = decrypt(wrapped_secret, key: wrapping_key) passkey_private_key = decrypt(WebAuthnCredentialSpecifics.encrypted_private_key, key:security_domain_secret) flags = { flag_user_present = 1, flag_user_present = 1 if user-verified else 0, flag_backup_eligible = 1, flag_backed_up_state = 1, } signature_counter = 0 rpId_hash = SHA_256(rpId) authenticator_data = {rpId_hash, flags, signature_counter} signed_data = authenticator_data + client_data_json assertion_signature = sign(signed_data, key:passkey_private_key) return to the client: AuthenticatorAssertionResponse { clientDataJSON: client_data_json, authenticatorData: authenticator_data signature: assertion_signature, userHandle: WebauthnCredentialSpecifics.credential_id } Chrome forwards the AuthenticatorAssertionResponse to the relying party, which verifies the signature using the previously registered passkey_public_key and authenticates the user. All requests sent to the cloud authenticator, including device management, key handling, recovery operations and passkey creation or use, are protected by a secure communication protocol. For example, once a WebAuthn API is issued and the user selects GPM as the passkey provider, Chrome initiates secure communication with the cloud authenticator as shown in Figure 9. Chrome uses a Google OAuth2 access token as the primary authorization signal for cloud authenticator operations. This token is issued for the Google account that is currently signed in. The token includes a dedicated scope: hxxps[:]//www.googleapis[.]com/auth/secureidentity.action. To obtain the token, Chrome exchanges a locally stored refresh token for a short-lived access token using Google’s OAuth2, as shown in Figure 10. Once the token is obtained, Chrome opens a WebSocket connection to the cloud authenticator. See Figure 11, wss[:]//enclave.ua5v[.]com/enclave. The cloud authenticator returns a WebSocket upgrade response (101 Switching Protocols), and Chrome proceeds to the Noise-NK handshake. Chrome and the cloud authenticator establish an encrypted session using the Noise Protocol Framework. Noise is a framework for flexible cryptographic handshakes that specifies a protocol for two parties to exchange Diffie-Hellman (DH) public keys. It then hashes the DH results into a shared secret and derives symmetric keys to protect all subsequent messages. Chrome uses the following handshake variant: Noise_NK_P256_AESGCM_SHA256. This combination defines: NK: the handshake pattern - N: the initiator (Chrome client) is unauthenticated - K: the responder (cloud authenticator) has a known static public key (the key is hard-coded in Chrome) P256: DH uses the NIST P-256 elliptic curve. AESGCM: encryption uses AES-GCM. SHA256: hashing uses SHA256. The session begins in an initial handshake state, where both sides prepare to exchange ephemeral keys and progressively mix cryptographic material into the handshake hash state and shared encryption key. Chrome sends the first handshake message, which includes its ephemeral public key and the result of an Elliptic Curve Diffie-Hellman (ECDH) operation with the cloud authenticator’s static public key. The cloud authenticator responds with its own ephemeral public key and performs a second ECDH operation using Chrome’s ephemeral key. To this message, the cloud authenticator also attaches an attestation signature for its Oak execution environment, intended to allow the client to verify that it is communicating with a trusted authenticator. We did not observe where, or if, Chrome validates this attestation. After the handshake messages are exchanged, Chrome and the cloud authenticator share a symmetric transport key and a handshake hash. Chrome can then send requests over the secure tunnel, each signed with a device key bound to the handshake. The device-to-cloud request is signed with one of the device’s hardware-backed keys. (The only exception is the initial onboarding message, which carries the device public keys.) For each request, Chrome determines which device key to use based on the context. When WebAuthn’s user verification is required or preferred, Chrome signs the request with the UV key, prompting the local user to verify before signing. When user verification is discouraged, Chrome uses the identity key instead. To bind each request to the active Noise session, Chrome creates a signature over both a serialized Concise Binary Object Representation (CBOR)-encoded request and the handshake hash. The signature not only proves the device's hardware identity and the integrity of the requested message, but also binds it to the current encrypted session. Below is an example using a passkeys/assert request. (Other device-to-cloud authenticator requests follow a similar structure, with the request fields changed.) requests = { "cmd": "passkeys/assert", "request": {rpId, challenge, userVerification,..,} "protobuf": WebauthnCredentialSpecifics, "wrapped_secret": wrapped_secret, "client_data_json": clientDataJSON } serialized_requests = CBOR.encode([requests]) serialized_requests_hash = sha256(serialized_requests) to_sign_message = handshake_hash || serialized_requests_hash Using Windows TPM-Backed Keys for the Signature For the identity key, Chrome signs the message using the Windows Cryptography Next Generation (CNG) APIs: NCryptOpenStorageProvider: Opens the MS_PLATFORM_CRYPTO_PROVIDER NCryptImportKey: Imports and unseals the wrapped_identity_private_key within the TPM NCryptSignHash: produces the signature on the message using the imported key When Chrome needs to sign with the UV key, it calls RequestSignAsync using the UV key label loaded from the passkey_enclave_state file. Windows Hello handles the user verification step, and after approval, Windows performs the actual signing inside the TPM and returns the resulting signature. With the signature ready, Chrome appends it to the request, encrypts it with the shared transport key and sends it over the WebSocket tunnel. request_body_map = { "sig": signature, "device_id": device_id, "auth_level": "uv", // or "hw" tag if signed with identity key "encoded_requests": serialized_requests } encode_message = CBOR.encode(request_body_map) // Make the final message length a multiple of 32 bytes message = encode_message || zero padding || pad_length_byte encypted_massage = noise.encrypt(message) websocket.send(encypted_message) The cloud authenticator decrypts the message using the shared transport key and verifies the signature using the stored device public key associated with the device_id. After processing the request, it prepares a CBOR-encoded response and encrypts it with the same transport key, and then sends it back to the client. Google Cloud Authenticator marks a fundamental shift in how passkeys are created, protected and used across devices. Passwordless authentication has traditionally followed two distinct paths: Hardware-bound keys, which offer strong protection but are locked to a single device Software-based keys, which sync easily but are far more vulnerable to theft The cloud authenticator introduces a new hybrid model. Sensitive key operations are moved to an isolated cloud environment. Every request remains anchored to hardware-backed keys on the user’s device. This approach allegedly preserves hardware-level assurances while enabling the global usability needed for seamless, synchronized cross-device authentication and recovery. This analysis lays the groundwork for highlighting the strengths of cloud-based authenticators. It sets the stage for an upcoming third post, where we’ll explore the new attack vectors in passwordless authentication. This includes cloud-based weaknesses that could allow a remote attacker to impersonate an existing synced device and obtain valid passkey authentication. Palo Alto Networks customers are better protected from threats that take advantage of issues with cloud authentication through the following products and services: Cortex Cloud Identity Security encompasses Cloud Infrastructure Entitlement Management (CIEM), Identity Security Posture Management (ISPM), Data Access Governance (DAG) and Identity Threat Detection and Response (ITDR). It provides clients with the necessary capabilities to improve their identity-related security requirements. By providing Cortex Cloud visibility into which identities are adding devices or altering their permissions within cloud environments, Cortex Cloud can accurately detect misconfigurations, unwanted access to sensitive data and real-time analysis surrounding usage and access patterns. CyberArk Identity Protection continuously maps authentication configurations and access posture across your human identity environment. It surfaces risks that passwordless deployments can obscure: accounts missing phishing-resistant MFA, misconfigured OAuth token lifetimes, dormant identities with persistent access, and privilege gaps lacking just-in-time controls. Its threat detection capabilities monitor for anomalous authentication patterns and suspicious access behaviors in real time. This enables security teams to identify and respond to identity-based attacks before they escalate, including cases where the initial access vector exploits trusted authentication flows like passkeys or synced credentials. Unit 42 Cloud Security Assessment is an evaluation service that reviews cloud infrastructure to identify misconfigurations and security gaps. If you think you may have been compromised or have an urgent matter, get in touch with the Unit 42 Incident Response team or call: North America: Toll Free: +1 (866) 486-4842 (866.4.UNIT42) UK: +44.20.3743.3660 Europe and Middle East: +31.20.299.3130 Asia: +65.6983.8730 Japan: +81.50.1790.0200 Australia: +61.2.4062.7950 India: 000 800 050 45107 South Korea: +82.080.467.8774 Palo Alto Networks has shared these findings with our fellow Cyber Threat Alliance (CTA) members. CTA members use this intelligence to rapidly deploy protections to their customers and to systematically disrupt malicious cyber actors. Learn more about the Cyber Threat Alliance.

AppViewX acquires Eos to extend identity security to AI agents and workloads AppViewX has acquired Eos, an AI-native identity control plane for AI agents and autonomous workloads within the enterprise. By combining AppViewX’s automated CLM and PKI with Eos’s agentic governance and privileged access control, the platform delivers an integrated solution for AI agent and machine identity security. The acquisition accelerates AppViewX’s evolution into an AI-native platform, positioning the company as a leader in securing machines, workloads, and AI agents in cloud and hybrid environments. Additionally, Archit Lohokare, CEO of Eos, has been appointed CEO of AppViewX, succeeding Dino DiMarino, who has stepped down as the company’s CEO to pursue other opportunities. Mr. Lohokare brings more than two decades of experience in cybersecurity across identity security, data security, and endpoint security. Prior to founding Eos, Mr. Lohokare was the SVP and GM of the workforce and endpoint security business at CyberArk, where he was instrumental in creating CyberArk’s Identity Security Platform. His expertise in identity security will guide AppViewX through its next phase of platform expansion in the AI and quantum era. “AI agents are increasingly acting with autonomy inside the enterprise, with privileged access to data, applications, infrastructure, and cloud environments,” said Archit Lohokare, CEO of AppViewX. “Identity is the control plane for this new era. As organizations rapidly deploy AI agents, they need a unified governance and risk layer that secures machines and AI agents with consistent visibility, policy, and enforcement. By combining AppViewX’s leadership in machine identity with Eos’s AI-native capabilities, we are building the next-generation platform enterprises need to innovate confidently and securely at AI scale.” Additionally, Kashyap Ivaturi, CTO of Eos, will be joining AppViewX as CTO and Madhu Venkatarajan, current AppViewX CTO, will be assuming the role of Chief Engineering Officer and GM of India Operations, focusing on scaling engineering excellence, accelerating product delivery, and strengthening AppViewX’s India-based innovation and execution capabilities. Mr. Ivaturi has led large-scale global enterprise engineering organizations at CyberArk and ServiceNow and brings extensive experience building secure, scalable SaaS platforms for global enterprises. His leadership will accelerate AppViewX’s product innovation and AI-native roadmap. “As organizations scale AI agents and autonomous workloads, the identity-driven attack surface expands dramatically,” said Ian Loring, Senior Managing Director & Executive Chair at Haveli Investments. “AppViewX has established leadership in machine identity management, and the addition of Eos accelerates its evolution into a unified platform that governs machines, AI workloads, and autonomous agents. We are confident that this combination, in addition to Archit’s and Kashyap’s leadership of the new company, will position AppViewX for long-term category leadership and durable growth.” “AI agents are a new identity group, and their rapid proliferation creates new risks for the enterprise,” said John Barrow, CISO at JB Poindexter. “These agents often behave non-deterministically. To reduce risk, we must monitor, audit, and control their privileged access to sensitive data and systems and ensure their lifecycle is automated and governed. I’m excited to see AppViewX and Eos come together. Eos’s AI agent Identity Security platform combined with AppViewX’s Machine Identity Security platform positions AppViewX as the leader in Machine and AI Agent Identity Security.” The acquisition of Eos adds to AppViewX’s growing market momentum and leadership recognition from both the IDC MarketScape for Certificate Lifecycle Management (CLM) and KuppingerCole’s Non-Human Identity Management (NHIM) Leadership Compass.

A new report from Grip Security analyzes 23,000 SaaS application environments. The statistics it found include: 100% of analyzed companies operate SaaS environments with embedded AI; there has been a year-over-year 490% spike in public SaaS attacks; and 80% of documented incidents involve PII and/or customer data. “But what really surprised me,” says Chad Holmes, product marketing consultant at Grip Security, “is that organizations have an average of 140 AI-enabled SaaS environments.” If an AI-enabled app is breached, any integral agentic AI can be used first to access data from connected systems, secondly to cascade from the one breach to a breach of every other AI-enabled environment within the organization – and potentially to expand further into AI-enabled environments in other organizations. This is chaos. The poster boy example of this cascading chaos is the Salesloft Drift incident (the ‘Great SaaS Breach of 2025’). Ultimately more than 700 organizations were affected, including security firms Cloudflare, Palo Alto Networks, Zscaler and CyberArk. UNC6395 attackers compromised Salesloft’s internal systems, starting with their GitHub repositories and moving from there into the Drift AWS environment. Here they stole the active OAuth and refresh tokens used by customers to connect the Drift Chatbot to local installations of Salesforce and other apps such as Slack. Armed with the legitimate pre-approved OAuth token, the attackers were able to impersonate Drift and log directly into Salesforce installations into companies also using the Drift chatbot. One breach of a SaaS app (Drift) cascaded into hundreds of compromises in different companies across the globe. Now it is fair to say that this incident prompted a drive to improve the security of SaaS apps and their AI implementations, but Grip is not convinced it will be enough. “One thing we’re seeing,” comments Holmes, “as we’ve moved outside the traditional perimeter, outside firewalls and network level protections, identity is the new perimeter. The focus is on identity, and if we have that identity, we can log into any environment anywhere.” For attacks against SaaS AI, the key ‘identity’ is a valid OAuth token. It’s worth briefly examining why this is such a threat, if control hasn’t been achieved. Much of it is down to the current need for speed in business. SaaS developers are tempted to rapidly include agentic AI within their own products to improve efficiency ahead of their competition, and they don’t always make the implications apparent to their customers. The customer may have installed shadow AI without knowledge. ‘Shadow’ is the epithet used to describe the use of AI or autonomous agentic AI without formal oversight from the IT and security departments – and if the customer is unaware of the AI within the SaaS app, it is automatically ‘shadow’. Customers adopt these apps too hastily, again to rapidly improve their own efficiency, and often without auditing them. Meanwhile, they have become so accustomed to issuing OAuth tokens that they may do so automatically as required by the SaaS app without considering wider implications from installing shadow AI. While complexity is the enemy of security, SaaS is the disguiser and multiplier of complexity, through poor visibility into its shadow AI. An attacker can often find greater visibility into a SaaS app by stealing the right OAuth access and/or refresh token (courtesy of the modern infostealer that can enter, scrape and depart without the victim realizing it). Armed with the right OAuth token, the attacker can enter the app unhindered, and start gathering data from whatever other systems are connected to the agentic system simply by feeding it tailored prompts through the APIs. However, the danger is not limited to the single SaaS app containing the shadow AI’s own environment – it can cascade to other environments within the organization. IdentityMesh can be an unintentional flaw resulting in a single unified authentication context. This allows the attacker inside the SaaS app to access all other agentic systems within the environment. If one of these has access to third party apps or shared service accounts used across organizations, the attacker can then pivot the attack across multiple organizations – and the cascading effect caused by a single stolen OAuth token in one company can cause the loss of data in many others. “AI is not a future risk, nor is it ‘just an IT problem’,” notes the report. “And crucially, governing it is not optional. It is now one of the most influential forces shaping how modern businesses operate and take on risk.” The report suggests that 2026 may be the worst year yet for SaaS breaches, warning that the increased blast radius may expand further as autonomous workflows outpace existing security controls. There are attempts at regulation, but globally, such regulations are moving in different directions with conflicting mandates, uneven enforcement and increased compliance friction. “AI regulation will get messier before it gets clearer,” says the report. But the chaos can be brought under control. “The way out is not more policy or slower innovation. It is a shift in how AI is governed in practice,” adds the report. The key is increased visibility into, and understanding of, SaaS shadow AI, and more dynamic governance. “Leaders who succeed replace static approvals with continuous oversight, discovery, and risk-based controls. AI becomes a managed third-party risk, monitored continuously, aligned to business outcomes, and governed with the same rigor as any critical supplier.” Related: The Blast Radius Problem: Stolen Credentials Are Weaponizing Agentic AI Related: Microsoft Highlights Security Risks Introduced by New Agentic AI Feature Related: Infostealers: The Silent Smash-and-Grab Driving Modern Cybercrime

Unit 42 is tracking an increased risk of wiper attacks related to the conflict with Iran, including multiple related incidents impacting organizations in Israel and the US. For the latest intelligence on cyberattacks associated with this conflict, review our Threat Brief: March 2026 Escalation of Cyber Risk Related to Iran. The primary vector for recent destructive operations from the Handala Hack group (aka Void Manticore, COBALT MYSTIQUE and Storm-1084/Storm-0842) reportedly involves the exploitation of identity through phishing and administrative access through Microsoft Intune. Handala Hack first emerged in late 2023. Despite initial hacktivist-aligned messaging, the group is currently assessed by the threat intelligence community to be a state-directed front for Iran’s Ministry of Intelligence and Security (MOIS). On March 6, Israel’s National Cyber Directorate warned of Iranian cyberattacks targeting Israeli organizations with wipers: “The National Cyber Command has received reports of several cases in which attackers gained access to corporate networks and deleted servers and workstations, with the aim of disrupting the operations of the attacked organizations. In some cases, the attacker had access data from legitimate corporate users, which was used to gain initial access to the network.” — Translated from source: Israel’s National Cyber Directorate. The following recommendations are based on the information reported publicly so far and threat intelligence from Palo Alto Networks Unit 42, specifically addressing the tactics observed by the Iranian-linked threat actor Handala. Persistent administrative rights are the single greatest risk factor in modern identity attacks. Attackers such as Handala target high-value accounts with "standing" (always-on) permissions to facilitate immediate impact. Just-in-time (JIT) access: Implement a JIT model for all administrative roles. Credentials should have zero permissions by default and only gain elevated rights through a formal activation process. A cloud infrastructure and identity management (CIEM) solution can help pinpoint identity risk in cloud resources. Microsoft Entra Privileged Identity Management (PIM): Use Entra ID PIM to manage eligible role assignments. Require multi-factor authentication (MFA), business justification and, for high-risk roles, manual approval before activation. CyberArk Privileged Access Management (PAM): For organizations with hybrid or complex multi-cloud environments, use CyberArk to vault administrative credentials and manage session isolation. CyberArk can provide a secure landing zone for administrators, designed to ensure that credentials for platforms like Intune never reside on a potentially compromised endpoint. Limit count: Reduce the number of Global Administrator and Intune Administrator accounts to the fewest possible based on business needs. A tool like the Cortex Identity Security dashboard can help discover which identities hold administrative privileges. Cloud-native accounts: Use cloud-only accounts (e.g., [email protected]) for administrative roles to prevent lateral movement from on-premises Active Directory via synchronized account compromise. Break-glass accounts: Maintain two emergency-access accounts that are excluded from standard conditional access policies, but protected by hardware-based MFA and monitored with high-severity alerts. Consider allowing mass wipe capabilities only from break-glass accounts. Enable multi-administrator approval (MAA): MAA requires a second, different administrator to review and approve high-impact actions before they are executed. Create an access policy for actions like wipe or delete. Role-based access control (RBAC): Use the Intune Administrator role specifically, rather than granting Global Administrator rights to device management staff. Inventory Service Principals with permissions for device management such as DeviceManagementManagedDevices.ReadWrite.All. PIM for Groups: Instead of assigning roles to individuals, use PIM for Groups (formerly Privileged Access Groups). Assign the Intune Administrator role to a security group and make users Eligible for membership in that group. This allows for unified auditing and approval workflows. Conditional access for elevation: Enforce authentication strength policies during PIM activation. Require FIDO2 hardware keys (YubiKeys) or Windows Hello for Business to activate roles that have the power to issue wipe commands. And allow sign-ins only from corporate IP address ranges or trusted locations. Leverage Secure Administrative Workstations (SAWs) and require Global Administrators to access Azure from hardened Privileged Access Workstations (PAWs). Leverage dedicated machines used only for administrative and sensitive data handling activities. Use enforced endpoint compliance before access is allowed. Reduce session lifetimes: Shorten session duration for sensitive administrative portals (e.g., Intune, Entra and Azure portals) to under 1 hour. This helps limit the area of impact for a stolen session token. Token Protection: Enable Token Protection (currently in preview for Entra ID) to cryptographically bind session tokens to the specific device from which they were issued, to help prevent an attacker from replaying them on a different machine. Tools such as the Cortex XDR authentication bypass module can help protect against attacks that attempt to circumvent authentication controls such as tokens. Discover and label sensitive data: Use data security posture management (DSPM) capabilities to scan and label sensitive data in the corporate hybrid environment. This classification enables granular segmentation, persistent encryption and automated security controls. Doing so helps ensure the organization’s most critical assets are protected regardless of where they reside. Leverage data loss prevention (DLP): Implement technologies such as the Palo Alto Networks AI-powered Enterprise DLP to alert and proactively block data exfiltration attempts. If storage accounts send significantly more data outbound than usual, organizations should immediately investigate. Managed detection and response (MDR)/extended detection and response (XDR) integration: Ensure audit logs (specifically RemoteWipe and FactoryReset actions) from device management tools such as Intune, are ingested into your security information and event management (SIEM)/XDR platform. Leverage automation, such as a security orchestration, automation and response (SOAR) platform, to rapidly respond to malicious events. A SOC platform such as Cortex XSIAM can perform these functions within one solution. Anomalous activity alerts: Configure specific alerts for mass wipe events. If more than a specific threshold of devices (e.g., five or 10) is targeted for a wipe within a short window, the system should trigger an immediate automated lockout of the initiating administrator account. Monitor Entra sign-in logs that would allow for detections and alerting if an administrator signs in from a different location (such as signing in from a new country) or outside of approved networks. Offline backups: Maintain immutable, air-gapped, offline backups of critical data. As the threat actor’s goal is often pure disruption (wiper activity) rather than financial extortion, the ability to restore from an immutable source may be the only guarantee of recovery. End-user training and tabletop exercises: Perform frequent phishing exercises, conduct staff cybersecurity training and hold tabletop exercises focused on destructive threat actor activities. If you think you may have been compromised or have an urgent matter, get in touch with the Unit 42 Incident Response team or call: North America: Toll Free: +1 (866) 486-4842 (866.4.UNIT42) UK: +44.20.3743.3660 Europe and Middle East: +31.20.299.3130 Asia: +65.6983.8730 Japan: +81.50.1790.0200 Australia: +61.2.4062.7950 India: 000 800 050 45107 South Korea: +82.080.467.8774 Additional Resources Threat Brief: March 2026 Escalation of Cyber Risk Related to Iran – Unit 42, Palo Alto Networks Iranian Cyber Threat Evolution: From MBR Wipers to Identity Weaponization – Unit 42, Palo Alto Networks Boggy Serpens Threat Assessment – Unit 42, Palo Alto Networks Updated March 13, 2026, at 1:05 p.m. PT to add links to resources. Updated March 23, 2026, at 3:33 p.m. PT, to add an Additional Resources section with links.

As the cybersecurity market continues to mature, SecurityWeek’s annual M&A report provides an essential roadmap for understanding how corporate consolidation is reshaping the digital defense sector. SecurityWeek’s M&A tracker cataloged 426 deals in 2025, including 334 involving pure-play companies. The 2025 figure of 426 represents a 5% increase over 2024, signaling that the sector is once again on an upward trajectory after two years of consecutive decline. While the total number of deals is lower than the 2022 peak (455), the value and complexity of the deals appear to be higher. In addition, the market continues to move away from the ‘buying everything’ trend observed in 2021/2022 to a more strategic, international consolidation. Regional dynamics and global footprint The 2025 data shows that the market has become significantly more globalized. At 288 participations for 426 deals, the US remains involved in roughly 67% of all global cybersecurity M&A. This is a slight recovery from the 63% low in 2023, but still well below the 80% dominance seen in 2021. The UK saw a decline from 70 deals (2021) to 48 (2023). However, it rebounded to 67 in 2024 and maintained that momentum with 64 deals in 2025. The UK has solidified its position as the clear #2 global hub for cybersecurity M&A. After being overtaken in the rankings in 2024, Israel has reclaimed the 3rd position with 34 deals in 2025. Ireland and Sweden saw fewer than 10 deals in 2024 after a strong 2023. In 2025, they bounced back to 12 and 11 deals respectively. Capital allocation and valuation trends Financial details were disclosed for 74 deals for a total disclosed value of $92.5 billion. This includes 63 deals involving pure-play cybersecurity companies, with a total disclosed value of $84 billion. The total disclosed deal value has surged by approximately 82% year-over-year, which can largely be attributed to Google’s planned acquisition of cloud security giant Wiz for $32 billion. The industry has sustained its appetite for billion-dollar acquisitions, with 11 deals exceeding the $1 billion mark. In addition to Google-Wiz, seven deals involving pure-play cybersecurity firms surpassed $1 billion, including Palo Alto Networks acquiring CyberArk ($25B) and Chronosphere ($3.3B), ServiceNow’s acquisition of Armis ($7.7B) and Veza ($1B), Francisco Partners buying Jamf ($2.2B), Veeam Software acquiring Securiti AI ($1.7B), and Proofpoint acquiring Hornetsecurity ($1.8B). This suggests that large-scale platform consolidation is now a permanent fixture of the landscape. The $100M-$999M range was the primary engine of growth in 2025. This indicates that acquirers are focusing on mature, established cybersecurity companies rather than early-stage startups. This also suggests that the industry is currently in a phase of strategic scaling, where medium-sized specialized leaders are being swallowed by larger platforms to create all-in-one security suites. There has been a noticeable decline in the number of disclosed deals under $100M (down to 26 in 2025 from 32 in 2024). This implies that ‘tuck-in’ acquisitions of very small startups are either becoming less frequent or are less likely to have their financial terms disclosed compared to larger, more impactful deals. Sector deep-dive and domain expertise SecurityWeek data shows that 125 deals involved managed security solutions providers (MSSPs) in 2025, including 60 pure cybersecurity providers. This includes product distributors and companies that offer other products and services in addition to cybersecurity. While activity involving MSSPs has increased from 119 deals in 2024, it remains significantly lower than the 150+ seen in 2022 and 2023. After a sharp drop in 2024, the 2025 figure suggests the market has found its new normal. It is no longer in a freefall, but it hasn’t returned to the frantic buying spree of 2022-2023. The strategic value remains high as managed services continue to be the primary way small-to-medium enterprises consume security. While it’s important to keep track of MSSP deals, we are tracking them separately in an effort to get a better view of the other categories. Several core security sub-sectors experienced significant upward momentum in 2025, as acquirers seem to prioritize data sovereignty, identity integrity, and regulatory compliance. With 82 deals, GRC (governance, risk, and compliance) reached a five-year peak in 2025. While it has always been a top category (averaging 65 deals from 2021-2024), the jump to 82 suggests that companies are prioritizing regulatory requirements over purely technical defenses. Data protection recorded aggressive growth, surging from 44 deals in 2024 to 63 deals in 2025. This reflects an industry shift toward data-centric security. While still a smaller category, AI security is showing a clear upward trajectory. It recorded 8 deals in 2024 and has increased to 13 deals in 2025, mirroring the broader tech industry’s focus on securing LLMs and AI pipelines. After a significant slump in 2024 (dropping to 28 deals), identity has bounced back to 43 deals in 2025, returning to its historical levels from 2022 and 2023. While several sectors saw record growth, four categories experienced a cooling in M&A activity, reflecting a pivot away from reactive tools toward long-term governance and data resilience. One of them is incident response, which has seen a notable decline, dropping from a peak of 38 deals in 2024 to 25 deals in 2025. Private equity-led acquisitions have also cooled significantly. They spiked to 37 in 2023 but dropped to 18 in 2025, returning to the levels seen in 2021 and 2022. After a strong 2024 (16 deals), interest in companies specializing in industrial/OT cybersecurity appears to have leveled off, dropping back to 9 deals, the same number recorded in 2023. Application security saw a steady rise in M&A deals through 2024 (reaching 31), but dipped slightly to 26 in 2025. Two types of companies stand out for remaining steady throughout the years. Network security remains the most consistent category in the dataset. It has hovered almost exactly at the same level for four years, with roughly 40 deals announced every year between 2022 and 2025. The government contractors segment is also remarkably stable, consistently recording between 36 and 38 deals annually (except for a slight outlier of 43 in 2021). It sits at 36 deals for 2025. The 2025 data indicates a market shift away from reactive tools (incident response) and toward resilience and compliance infrastructure (GRC and data protection). While 2024 was a year of diversifying into specialized niches, 2025 appears to be a year of consolidation around the core pillars of governance and data privacy. Monthly summaries of 2025 cybersecurity M&A deals: January, February, March, April, May, June, July, August, September, October, November, December. Methodology: The data was collected through news distribution services, Google searches and pitches from PR companies. The data includes companies that issued press releases announcing or mentioning acquisitions, as well as deals that have been privately reported to SecurityWeek. All deals that had a cybersecurity component have been taken into account for this study. Mergers and acquisitions that did not have an English-language announcement may not be included. The data could also include some deals that may have not been completed after they were announced. The GRC category includes governance, compliance, risk management, audit, assessment, vulnerability management, penetration testing, attack surface management, offensive security, and cyber insurance. Network security includes endpoint security, MDR, XDR, NDR, and SASE. Identity includes IAM, PAM, secure access, authentication, and authorization. Incident response includes SOAR, SIEM, SOC, and forensics. ‘Other (specialized)’ includes hardware, quantum, media and events, healthcare, brand protection, financial, online safety, SAP, recruiting, real estate, and automotive. Data protection includes encryption/cryptography, VPN, privacy, backup and blockchain. MSSP includes cybersecurity solution distributors and companies that do not develop their own products or solutions. The AI category only includes companies specializing in securing AI.

Palo Alto Networks announced on Tuesday that it has entered into a definitive agreement to acquire endpoint security company Koi. Financial details have not been disclosed by the security firms, but Israeli publications Calcalist and Globes both reported that the deal is valued at $400 million. Koi, which raised $48 million in funding, offers an endpoint security platform that focuses on protecting various types of software, including applications, code, operating system packages, extensions, AI models, AI agents, and containers. The platform includes tools for automatic discovery and inventory of software across IT setups. It employs an AI-driven risk engine to analyze factors like code changes, runtime actions, update paths, and network outflows for real-time threat detection. Remediation features enable actions such as quarantining risky elements, reverting versions, and notifying owners. A supply chain gateway component curates incoming software from platforms like GitHub and Hugging Face. Palo Alto Networks said it’s buying Koi for its agentic endpoint security technology, aiming to enhance Palo Alto’s Prisma AIRS AI security platform and Cortex XDR endpoint security solution to provide “significant visibility into the AI attack surface to improve security policy and malware prevention”. “AI agents and tools are the ultimate insiders. They have full access to your systems and data, but operate entirely outside the view of traditional security controls,” said Lee Klarich, chief product and technology officer at Palo Alto Networks. “By acquiring Koi, we will be closing this gap and setting a new standard for endpoint security. We will give our customers the visibility and control required to safely harness the power of AI—ensuring that every agent, plugin, and script is governed, verified, and secure.” Koi has been in the news in recent months for its analysis of attacks and vulnerabilities impacting software developers. Palo Alto Networks announced three major acquisitions last year: observability platform Chronosphere for $3.35 billion, identity security giant CyberArk for $25 billion, and AI security firm Protect AI reportedly for over $500 million. SecurityWeek’s M&A tracker cataloged more than 420 deals in 2025. A detailed report will be published in the coming days. Related: EU Unconditionally Approves Google’s $32B Acquisition of Wiz Related: Cybersecurity M&A Roundup: 34 Deals Announced in January 2026 Related: Check Point Announces Trio of Acquisitions Amid Solid 2025 Earnings Beat

Betti RHC, la prima graphic novel al mondo dedicata alla cybersecurity awareness, ha finalmente il suo sito ufficiale. Uno spazio tutto suo dove scoprire il progetto, sfogliare le copertine degli episodi e immergersi nel mondo di Betti: la giovane laureanda in informatica che, dopo la morte misteriosa del padre, si trasforma nell'hacker più potente del mondo. Una storia avvincente che, episodio dopo episodio, affronta una minaccia digitale diversa — dal phishing al ransomware, fino al cyberbullismo — e insegna a riconoscerla e a difendersi, senza che sembri mai una lezione. Sul sito trovate tutto ciò che rende Betti un progetto diverso dal solito: la sua filosofia, le anteprime delle tavole e il racconto di come nasce ogni volume. Perché dietro Betti RHC c'è solo lavoro umano: ogni tavola è disegnata interamente a mano dagli artisti del Gruppo Arte di Red Hot Cyber, senza alcun uso di intelligenza artificiale. E a garantire che ogni storia sia realistica e tecnicamente corretta c'è la supervisione degli hacker etici del gruppo HackerHood, che mantengono il racconto fedele al mondo reale della sicurezza informatica. C'è spazio anche per le aziende, che possono usare Betti come strumento di awareness diverso dai soliti corsi: acquistare i volumi, personalizzarli con il proprio brand o sponsorizzare nuovi episodi. E come primo regalo, l'episodio "Byte the Silence", dedicato al cyberbullismo, è scaricabile gratuitamente per uso personale. Perché la miglior difesa, in fondo, è una bella storia. 👉 Scopri tutto su https://betti.redhotcyber.com/

Security researchers have detailed how they discovered and exploited a cross‑site scripting (XSS) vulnerability in a popular infostealer, allowing them to gather crucial evidence about its back-end operations. Ari Novick, a malware researcher at identity security specialist CyberArk, explained in a blog post that the XSS bug was found in the web panel of the StealC variant. “Given the core business of the StealC group involves cookie theft, you might expect the StealC developers to be cookie experts and to implement basic cookie security features, such httpOnly, to prevent researchers from stealing cookies via XSS,” he explained. “The irony is that an operation built around large-scale cookie theft failed to protect its own session cookies from a textbook attack.” The report focused on one StealC user, dubbed “YouTubeTA,” who managed to steal 390,000 passwords and over 30 million cookies via the malware. The XSS exploitation enabled Novick to identify certain characteristics on the threat actor’s computers, including its geolocation, and retrieve active session cookies. Novick claimed most of the victims were on YouTube looking for cracked versions of Adobe Photoshop and Adobe After Effects, when they unwittingly installed StealC. “YouTubeTA was likely using StealC to take over old YouTube accounts, which they then used to promote new samples of StealC,” he explained. By studying hardware fingerprinting, supported languages, time zones and IP addresses connected to the threat actor’s use of StealC, Novick was able to deduce several things about them. Notably, they are using an Apple Pro device with an M3 processor supporting both English and Russian language settings, are based in the Eastern European time zone and access the internet via Ukrainian ISP TRK Cable TV. Pros and Cons of Infostealers Malware-as-a-service (MaaS) operations like StealC offer powerful out-of-the-box capabilities for threat actors like YouTubeTA, who was able to compromise a large group of victims in just a few months, said CyberArk. “This is a clear demonstration of why many threat actors employ the MaaS model,” Novick continued. “By delegating much of the work to other groups, they can specialize and have a more significant impact, much like in traditional industries.” However, by relying on such tools, threat actors also expose themselves to the same software supply chain risks as legitimate businesses. “The StealC developers exhibited weaknesses in both their cookie security and panel code quality, allowing us to gather a great deal of data about their customers,” Novick concluded. “If this holds for other threat actors selling malware, researchers and law enforcement alike can leverage similar flaws to gain insights into, and perhaps even reveal the identities of, many malware operators.”

Palo Alto Networks (NASDAQ: PANW) announced on Wednesday that it has agreed to acquire observability platform provider Chronosphere in a deal valued at $3.35 billion. Chronosphere explains that its platform enables teams to “zero in on the data that’s most useful” and provides insights into every layer of their stack — from the infrastructure to the applications to the business. Palo Alto said the combined solution of Chronosphere’s observability platform with Palo Alto’s AgentiX will deploy AI agents on massive amounts of data monitored by Chronosphere’s platform to detect performance issues, autonomously investigate the root cause, and close the loop with agentic remediation. “The foundational requirement for every modern AI data center is constant uptime and resilience, which demands real-time, always-on observability delivered at the right cost,” said Nikesh Arora, Chairman and CEO at Palo Alto Networks. Under the terms of the agreement, Palo Alto Networks will acquire Chronosphere for a total consideration of $3.35 billion, to be paid in cash and replacement equity awards, subject to adjustments. Chronosphere reported annual recurring revenue (“ARR”) of more than $160 million as of the end of September 2025. The move to acquire Chronosphere is the latest of several acquisitions in recent years under Arora and follows a massive $25 billion deal to acquire CyberArk which was announced in July 2025. In April the cybersecurity giant said it was acquiring AI security company Protect AI in a deal reportedly valued at roughly $650-700 million.

Google and Wiz said the antitrust review initiated by the United States Department of Justice into their planned $32 billion acquisition has been cleared. The companies announced reaching an agreement on the terms of an acquisition in March 2025. News of a Justice Department antitrust review into Google’s planned acquisition of the cloud security giant came to light in mid-June. The goal of the probe was to determine whether the deal would harm competition in the cybersecurity market. During the recent WSJ Tech Live California event, Wiz CEO Assaf Rappaport confirmed that his company had cleared the regulatory hurdle, noting they are “still in the journey between signing and closing”. The DOJ recently signaled the clearance of the merger by issuing an early termination of its investigation, indicating it will not challenge the deal further. “We look forward to completing the review process in other jurisdictions,” a Google spokesperson told the media. “The acquisition of Wiz is expected to close in 2026, subject to customary closing conditions.” A possible acquisition of Wiz by Google was first announced in July 2024, when the deal was reportedly valued at $23 billion. Wiz walked away, saying that it had decided to pursue an IPO. The new $32 billion deal between the two companies is the biggest cybersecurity acquisition ever, followed by Palo Alto Networks’s deal to acquire CyberArk for $25 billion. Cisco acquired Splunk for $28 billion in 2024, but that was a hybrid security and observability play, rather than a pure cybersecurity deal. Wiz has developed solutions designed to find and address security flaws in enterprise cloud deployments. Google has long sought to strengthen its presence in the enterprise cybersecurity space and the addition of Wiz technologies to Google Cloud — alongside Mandiant — will help the tech giant establish an end-to-end security platform, covering proactive defense, threat intelligence, and incident response. Google said Wiz security products will remain accessible across all major cloud platforms, including AWS, Microsoft Azure, and Oracle Cloud. Additionally, customers will continue to have access to Wiz’s offerings through partner security solutions. [ Read: What’s Behind Google’s $32 Billion Wiz Acquisition? ] Agnidipta Sarkar, chief evangelist at zero trust microsegmentation company ColorTokens, commented, “This strategic collaboration between Wiz and Google Cloud will likely result in more integrated security solutions for clients in the cloud sector. The industry is expected to see a continuation of innovation in cloud security, thereby delivering increased value to clients. This acquisition may also provide substantial advantages for Google’s enterprise cloud business.” Karen Walker, CFO at cloud security provider Sysdig, a Wiz competitor, said some Wiz customers have concerns. “Organizations are increasingly building infrastructure across multiple cloud and on-prem environments. For Wiz customers who’ve come to expect cloud security independence, the path forward has become a lot murkier. Wiz customers operating in other clouds are suddenly at risk of becoming second-class citizens, facing higher costs and slower innovation,” Walker explained. “At its scale, Google will naturally align Wiz more tightly with its own platform and roadmap, and the other cloud providers are likely to be less open to sharing their roadmap with Wiz. This roadmap sharing between partners is so important for security companies to stay aligned with cloud innovation,” Walker added. Related: Jamf to Go Private Following $2.2 Billion Acquisition by Francisco Partners Related: Cybersecurity M&A Roundup: 45 Deals Announced in October 2025

CyberArk launches Secure AI Agents to safeguard privileged AI identities CyberArk is announcing the general availability of the CyberArk Secure AI Agents Solution, expanding the CyberArk Identity Security Platform with privilege controls to secure the rapidly growing class of AI agent identities. As organizations adopt AI agents to automate tasks and drive efficiency, these autonomous entities are emerging as a powerful and privileged new identity class. AI agents introduce new risks, including hallucinations, misuse, and potential takeover by malicious actors. These risks are amplified when agents require elevated privileges. The CyberArk Secure AI Agents Solution addresses these challenges by applying precise privilege controls to ensure AI agents have only the access they need, when they need it, and nothing more. This approach reduces risk, prevents unauthorized access, and enables organizations to scale AI-driven initiatives with confidence. “As organizations embrace AI agents, both builders and defenders must understand how identity-centric risks evolve when agents require elevated privileges,” said Matt Cohen, CEO, CyberArk. “Without strong discovery, robust privilege controls, and comprehensive lifecycle management, organizations risk losing visibility and opening the door to catastrophic agentic attacks. CyberArk uniquely secures the full spectrum of identities — humans, machines and AI agents — by applying the right level of privilege controls, enabling innovation while maintaining security and compliance.” Securing AI agents requires a privilege-focused approach According to new CyberArk CISO research, AI agent adoption is expected to reach 76% within three years, yet fewer than 10% of organizations have adequate security and privilege controls in place. The report, Securing Agentic AI: Identity as the Emerging Foundation for Defense, reveals that: Nearly 40% of enterprise financial institutions and software companies already have agentic AI in production. Fewer than one in ten organizations have deployed agentic security controls such as risk registries and dynamic authorization at scale. Two-thirds of CISOs in financial services and software rank agentic AI among their top three cybersecurity risks, with more than one-third citing it as their top concern. Most expect AI agent security to drive increased cybersecurity spending in the coming year. AI agents act with autonomy, reasoning, and access to sensitive systems, often requiring privileged permissions to execute their functions. Without proper oversight, these privileges can be misused or hijacked, leading to potentially severe business and regulatory consequences. Privilege controls across every identity The CyberArk Identity Security Platform delivers comprehensive privilege controls across the full spectrum of identities: human, machine, and AI. With the introduction of the Secure AI Agents Solution, these proven capabilities are extended to autonomous AI agents, applying the same principles of just-in-time access, least privilege, and continuous session monitoring that have defined CyberArk’s leadership in identity security. This unified, privilege-first approach ensures that every identity is governed, secured, and monitored with the same rigor, enabling innovation without compromising security or compliance. The CyberArk Secure AI Agents Solution will deliver: Comprehensive agent discovery: Automatically detect AI agents across SaaS, cloud, and developer environments, with enriched profiles including ownership, roles, and access rights. Secure agent access: Enforce strong authentication and least-privilege access, with zero standing privileges and agent activity auditing. Real-time threat detection: Continuously monitor for anomalies and unauthorized access, triggering automated alerts and rapid response. Lifecycle management and compliance: Govern AI agent from creation through decommission, supporting evolving regulatory requirements and audit readiness.

Keycard emerges from stealth with identity and access solution for AI agents Keycard emerged from stealth with its identity and access platform for AI agents that integrates with organizations’ existing user identity solutions. Keycard’s platform identifies AI agents, lets users assign task-based permissions and dynamically enforces policy while tracking all activity. With Keycard, organizations can deploy AI agents into production with complete trust, knowing they are only capable of performing the intended actions of their users and builders. “AI agents represent a once-in-a-generation shift, greater than the SaaS and cloud wave combined. But without trusted access controls, they can’t leave the lab. Keycard provides the guardrails that allow agents to act safely on behalf of people and businesses, unlocking the true potential of the agent economy,” said Ian Livingstone, CEO of Keycard. Keycard was co-founded by Ian Livingstone, Matthew Creager and Jared Hanson. Livingstone and Creager served as senior leaders at Snyk where they built the platform engineering and developer experience divisions as revenues grew from $30 million to $300 million. Hanson was the Chief Architect at Auth0 before joining Okta as a senior technical leader, and he created Passport.js, the most popular authentication framework for Node.js that is downloaded millions of times every week. At these companies, the trio witnessed how developers constantly struggled to connect services and applications together within and across organizations at enterprise scale, often leading to painful security incidents and significant product delays. This problem has only been exacerbated with AI agents. Today they spawn by the thousands and operate autonomously across systems and organizational boundaries. They need different permissions based on the task they are working on, which people and companies they are doing the task for, the resources they are accessing and who owns those resources. Existing solutions were not architected for AI agents. They were built for a world of static, human-driven point-and-click interactions and not the machine-driven dynamic, ephemeral and autonomous nature of AI agents. AI is expected to drive the creation of the greatest number of new identities with privileged and sensitive access according to CyberArk’s 2025 Identity Security Landscape. Keycard purpose-built its identity and access platform for AI agents. Its approach is dynamic, contextual and built for trust in an AI agent driven world: Keycard cryptographically proves who an agent is, who they act for and whether they are authorized backed by federated, standards-based protocols – preventing lock-in to a few centralized platforms while interoperating with leading agents from companies like Anthropic, Microsoft and OpenAI. Keycard replaces static secrets and API keys with dynamic, identity-bound and task-scoped tokens that enable the enforcement of policy that adapts to the changing trust landscape of users, systems and companies without any code changes. These tokens are tied to the agent’s attested workload identity, compute provider or device for end-to-end trust. Keycard enforces contextual access controls at runtime, giving users and companies the tools to express relationship and task-based policy using its identity-aware data model, enabling the immediate revocation of access with a single API call. Keycard operates at internet scale, designed for agentic performance expectations: dynamic, ephemeral, performance-obsessed, globally available infrastructure that models the complex mixed-identity and resource relationships of human and agent interactions, a fundamentally different design than existing solutions. Keycard provides developers with a set of SDKs they can use to build trusted agentic applications without needing to be identity or security experts, which gives security the context and feedback loops required to govern AI agents as they are developed instead of after they are in production. Keycard contributes to emerging standards including Model Context Protocol (MCP), WIMSE and OAuth extensions for agents and is the first production implementation with support for OAuth 2.1 Client ID Metadata Documents in MCP. Keycard raises $38 million to expand its AI identity platform Keycard also announced that it has raised $38 million in seed and Series A funding. Andreessen Horowitz and boldstart ventures co-led the $8 million seed round, and Acrew Capital led the $30 million Series A. Angels who participated include Ian Andrews (CRO at Groq), Ryan Carlson (president at Chainguard), Emilio Escobar (CISO at Datadog), Karl McGuinness (former Chief Product Architect at Okta) and Matias Woloski (co-founder and former CTO at Auth0). “This is the Auth0 moment for agent access. The agent ecosystem needs foundational authorization infrastructure. This team has the rare combination of infrastructure expertise and standards leadership required to build it,” said Zane Lackey, partner at Andreessen Horowitz. “Winning in agentic security takes a rare mix: build for developers and nail security from day one. Ian, Matthew and Jared did it before at Snyk and Auth0 and they’re doing it again with Keycard. We’re thrilled to back Ian and Matthew a second time (their first company was acquired by Snyk) alongside Jared. This is the team to define the identity and trust category for the agent era,” continued Ed Sim, general partner at boldstart ventures. “Trusted agents define the next chapter of computing. Developers will lead the way and need durable identity and access foundations. Keycard extends trust and control to the agent layer,” concluded Asad Khaliq, founding partner at Acrew Capital. Keycard plans to use the funding to continue to advance its identity and access platform and to expand its R&D team.

CyberArk unveils new capabilities to reduce risk across human and machine identities CyberArk announced new discovery and context capabilities across its Machine Identity Security portfolio. The enhancements enable security teams to find, understand and secure machine identities, spanning certificates, keys, secrets, workloads and more, reducing risk and simplifying compliance at scale. Machine identities outnumber human identities by an estimated 82 to 1, driven by increased AI adoption and cloud native growth. As a result, machine identity-related security incidents are on the rise, with 72% of security leaders reporting certificate-related outages and 50% experiencing security incidents or breaches from compromised machine identities, according to CyberArk research. Manual processes can no longer keep up, and organizations need an automated discovery and context-driven approach to stay ahead. “Implementing machine identity security programs has become increasingly complex as organizations grapple with shrinking certificate lifespans, the rise of AI agents, vault sprawl and vulnerable software supply chains. With these new discovery, context and remediation capabilities, customers gain the visibility and control they need to tame sprawl, enforce policy and secure their environments more efficiently,” said Kurt Sand, GM of Machine Identity Security at CyberArk. CyberArk’s expanded Machine Identity Security portfolio delivers centralized visibility, automated policy enforcement and context-driven insights to help organizations monitor and secure every machine identity, anywhere, across the enterprise. Key enhancements include: CyberArk Secrets Hub Discovery and context for HashiCorp Vault – Helps address critical vault sprawl challenges by providing visibility into dispersed HashiCorp Vault instances and ensuring enterprise-wide policy compliance without disrupting developer workflows. Risk management and remediation dashboard – Centralizes observability across secrets vaults and integrates third-party scanner data to identify high-risk areas, enabling organizations to prioritize remediation and track compliance progress. CyberArk Certificate Manager, SaaS CA/B Forum TLS Certificate (47-day) Dashboard – Provides real-time visibility into certificate expiration timelines, renewal projections and certificate authority usage to help organizations prepare for reduced TLS certificate lifespans (from 398 days today to 200 days in 2026, 100 days in 2027 and 47 days by 2029), allowing them to easily manage renewals and prevent outages. Code sign management, policy enforcement and deep DevOps integrations – Provide automated, policy-enforced code signing and governance alongside certificate lifecycle management to reduce infrastructure overhead, accelerate adoption and help ensure only trusted, compliant software is released. CyberArk SSH Manager for Machines New authorization and policy controls – Grant real-time authorization tracking and discovery for centralized visibility, risk reduction and audit compliance to help better manage SSH key sprawl and unmitigated access.

Salesforce has confirmed that it will not negotiate with or pay a ransom to the threat actors behind a massive wave of data theft attacks that impacted the company's customers this year. As first reported by Bloomberg, Salesforce emailed customers on Tuesday to say they would not be paying a ransom and warned that "credible threat intelligence" indicates the threat actors were planning to leak the stolen data. "I can confirm Salesforce will not engage, negotiate with, or pay any extortion demand," Salesforce also confirmed to BleepingComputer. This statement follows the launch of a data leak site by threat actors known as "Scattered Lapsus$ Hunters," who are attempting to extort 39 companies whose data was stolen from Salesforce. The website was located on the breachforums[.]hn domain, which is named after the notorious BreachForums website, a hacking forum known for selling and leaking stolen data. The companies being extorted on the data leak site included well-known brands and organizations, including FedEx, Disney/Hulu, Home Depot, Marriott, Google, Cisco, Toyota, Gap, Kering, McDonald's, Walgreens, Instacart, Cartier, Adidas, Sake Fifth Avenue, Air France & KLM, Transunion, HBO MAX, UPS, Chanel, and IKEA. In total, the threat actors claimed to have stolen nearly 1 billion data records, which would be publicly released if an extortion demand was paid by individual companies or as a single payment from Salesforce that would cover all the impacted customers listed on the site. This data was stolen from Salesforce instances in two separate campaigns that occurred in 2025. The first data theft campaign began at the end of 2024, when threat actors started conducting social engineering attacks impersonating IT support staff to trick employees into connecting a malicious OAuth application to their company's Salesforce instance. Once linked, the threat actors used the connection to download and steal the databases, which were then used to extort the company through email. These social engineering attacks impacted Google, Cisco, Qantas, Adidas, Allianz Life, Farmers Insurance, Workday, Kering, and LVMH subsidiaries, such as Dior, Louis Vuitton, and Tiffany & Co. A second Salesforce data-theft campaign began in early August 2025, when the threat actors used stolen SalesLoft Drift OAuth tokens to pivot to customers' CRM environments and exfiltrate data. The Salesloft data-theft attacks primarily focused on stealing support ticket data to scan for credentials, API tokens, authentication tokens, and other sensitive information that would enable the attackers to breach the company's infrastructure and cloud services. One of the threat actors behind the Salesloft attacks, known as ShinyHunters, told BleepingComputer that they stole approximately 1.5 billion data records for over 760 companies during this campaign. Many companies have already confirmed they were impacted by the Salesloft supply-chain attack, including Google, Cloudflare, Zscaler, Tenable, CyberArk, Elastic, BeyondTrust, Proofpoint, JFrog, Nutanix, Qualys, Rubrik, Cato Networks, Palo Alto Networks, and many more. The recently launched data leak site was used primarily to extort customers in the original social engineering attacks, with the threat actors stating they would begin publicly extorting those impacted by the Salesloft attacks after October 10th. However, the data leak site is now shut down, with the domain now using nameservers of surina.ns.cloudflare.com and hans.ns.cloudflare.com, which have both been used by the FBI in the past when seizing domains. BleepingComputer contacted the FBI about whether they seized the domain but has not received a response at this time. Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply. The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments. Get the report

An extortion group has launched a new data leak site to publicly extort dozens of companies impacted by a wave of Salesforce breaches, leaking samples of data stolen in the attacks. The threat actors responsible for these attacks claim to be part of the ShinyHunters, Scattered Spider, and Lapsus$ groups, collectively referring to themselves as "Scattered Lapsus$ Hunters." Today, they launched a new data leak site containing 39 companies impacted by the attacks. Each entry includes samples of data allegedly stolen from victims' Salesforce instances, and warns the victims to reach out to "prevent public disclosure" of their data before the October 10 deadline is reached. The companies being extorted on the data leak site include well-known brands and organizations, including FedEx, Disney/Hulu, Home Depot, Marriott, Google, Cisco, Toyota, Gap, McDonald's, Walgreens, Instacart, Cartier, Adidas, CarMax, Sake Fifth Avenue, Air France & KLM, Transunion, HBO MAX, UPS, Chanel, and IKEA. "All of them have been contacted long ago, they saw the email because I saw them download the samples multiple times. Most of them chose to not disclose and ignore," ShinyHunters told BleepingComputer. "We highly advise you proceed into the right decision, your organisation can prevent the release of this data, regain control over the situation and all operations remain stable as always. We highly recommend a decision-maker to get involved as we are presenting a clear and mutually beneficial opportunity to resolve this matter," they warned on the leak site. The threat actors also added a separate entry requesting that Salesforce pay a ransom to prevent all impacted customers' data (approximately 1 billion records containing personal information) from being leaked. "Should you comply, we will withdraw from any active or pending negotiation indiviually from your customers. Your customers will not be attacked again nor will they face a ransom from us again, should you pay," they added. The extortion group also threatened the company, stating that it would help law firms pursue civil and commercial lawsuits against Salesforce following the data breaches and warned that the company had also failed to protect customers' data as required by the European General Data Protection Regulation (GDPR). Scattered Lapsus$ Hunters have been targeting Salesforce customers with voice phishing attacks since the beginning of the year, leading to breaches that have impacted companies such as Google, Cisco, Qantas, Adidas, Allianz Life, Farmers Insurance, Workday, as well as LVMH subsidiaries, including Dior, Louis Vuitton, and Tiffany & Co. In these attacks, the threat actors tricked employees into linking a malicious OAuth app to their company's Salesforce instance. ShinyHunters told BleepingComputer that while a particular Salesforce instance may have been targeted, it also contained data for many of the subsidiaries, making the attacks more impactful. Once connected, the attackers stole company databases and used the data to extort victims via email. These extortion emails were signed by ShinyHunters, a notorious extortion group linked to a long string of high-profile breaches in recent years, including the Snowflake attacks and those against AT&T and PowerSchool. ShinyHuntersalso claimed to have used stolen OAuth tokens for Salesloft's Drift AI chat integration with Salesforce to steal sensitive information, including passwords, AWS access keys, and Snowflake tokens, from customers' Salesforce instances. These attacks were tracked by Mandiant under a separate threat cluster named "UNC6395" as they have been unable to formally link the breaches to this group. On a Telegram channel associated with the extortion group, the threat actors claim that they will begin extorting companies affected by the Salesloft Drift attacks on a separate data leak site launching on October 10th. The Salesloft attacks are known to have impacted Google, Palo Alto Networks, CyberArk, Cloudflare, Rubrik, Elastic, BeyondTrust, Proofpoint, JFrog, Zscaler, Tenable, Nutanix, Qualys, and Cato Networks, among many others. ShinyHunters claim that companies will not be re-extorted under the Salesloft campaign if a ransom is paid in this initial extortion phase. "We are aware of recent extortion attempts by threat actors, which we have investigated in partnership with external experts and authorities. Our findings indicate these attempts relate to past or unsubstantiated incidents, and we remain engaged with affected customers to provide support," Salesforce said in a statement published after ShinyHunters launched their data leak site. "At this time, there is no indication that the Salesforce platform has been compromised, nor is this activity related to any known vulnerability in our technology." Update October 03, 11:02 EDT: Added Salesforce statement. Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply. The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments. Get the report

Cybersecurity jobs available right now: September 30, 2025 Application Security Architect Vancity | Canada | Hybrid – No longer accepting applications As an Application Security Architect, you will lead the design and planning of information security projects, including cost estimation and forecasting. Contribute to the development of enterprise security documentation such as policies, standards, baselines, guidelines, and procedures. Assess security risks across programs, projects, and operational processes, and recommend architecture remediation strategies. CISO UVeye | Israel | Hybrid – No longer accepting applications As a CISO, you will develop and implement a comprehensive cybersecurity strategy and architecture ensuring the highest protection against cyber threats. Lead incident response efforts, including coordinating with internal teams and external partners, to mitigate security incidents promptly. Implement advanced threat detection and monitoring tools to identify and mitigate security vulnerabilities and potential breaches proactively. Cloud SecOps Lead RIB Software | India | Remote – View job details As a Cloud SecOps Lead, you will lead and develop the Security Operations team, defining processes for monitoring, response, and continuous improvement. Oversee SIEM and SOAR operations, ensuring effective detection, alert triage, and automated incident response. Coordinate incident response efforts end-to-end, including containment, remediation, and post-incident reviews. Get weekly updates on new cybersecurity job openings. Subscribe here! Cybersecurity Architect Voith Hydro | Germany | Remote – No longer accepting applications As a Cybersecurity Architect, you will develop cyber security concepts and architectures with the development teams to fulfill the respective security target levels supports, including threat and risk analysis. Drive the implementation of cyber security measures over the whole lifecycle of Voith Hydro products, and initiate penetration tests for high-risk areas for independent validation of issues and remediation efforts as well as analyze and evaluate specific customer project security requirements. Cybersecurity Engineer MANTECH | USA | On-site – No longer accepting applications As a Cybersecurity Engineer, you will conduct in-depth analysis of cyber security tools, applications, and methodology to maximize and optimize the enterprise cyber security posture using reverse engineering techniques to understand their behavior, functionality, and impact. Utilize tools and techniques providing advice to tune enterprise cyber security tools to scan, stop, and alert on threats and incidents enabling event/incident analysis, troubleshooting, and resolution. Cybersecurity Engineer Tenet Healthcare | USA | On-site – No longer accepting applications As a Cybersecurity Engineer, you will serve as the SME for the cybersecurity operations team and the main escalation point for SIEM issues, supporting root cause analysis and service restoration. You will help shape the enterprise-wide cybersecurity strategy by collaborating with stakeholders and Tenet service providers. Report security gaps to leadership with appropriate recommendations. Cyber Security Consultant PwC | India | On-site – No longer accepting applications As a Cyber Security Consultant, you will design and implement a secure OT architecture, including network segmentation, firewalls, access controls, and specific OT security measures. Oversee incident response activities for OT, including detection, containment, eradication, and recovery, ensuring business continuity and minimal disruption. Cyber Security Consultant Daintta | United Kingdom | Hybrid – No longer accepting applications As a Cyber Security Consultant, you will support the development and roll out of cyber security strategies, capabilities, and operating models to meet clients’ goals. Undertake risk and maturity assessments, providing strategic guidance on how to improve cybersecurity posture using a risk-based approach with articulation of mitigations/ controls and their respective impact on reducing risk. Cyber Security Engineer UNSW College | Australia | On-site – No longer accepting applications As a Cyber Security Engineer, you will monitor SIEM and other security platforms for alerts, triaging and escalating incidents as needed. You will execute first- and second-level response procedures, including containment, investigation, and documentation. You will administer and maintain endpoint protection, email security, firewalls, and other security infrastructure, run vulnerability scans, record findings, and coordinate remediation with the appropriate teams. Cyber Security Engineer CLUSTER REPLY | Italy | Hybrid – No longer accepting applications As a Cyber Security Engineer, you will design security architectures for cloud and hybrid environments based on Microsoft Azure and Microsoft 365. Implement protection solutions using Microsoft Defender, Sentinel, Azure Security Center, Purview, and Intune. Manage identity and access through Azure AD, MFA, PIM, and Conditional Access. Cyber Threat Investigator Accenture | United Kingdom | Hybrid – No longer accepting applications As a Cyber Threat Investigator, you will develop, maintain, and enhance security detection content primarily for the Splunk SIEM, to enable the detection of threats across diverse platforms. Perform security monitoring, reviewing and triaging triggered alerts, and suggesting improvements. Collaborate with the extended security team to identify gaps in detection coverage, log ingestion and alerting based on business risks and threats. Information Security – Firewall & Network Engineer Starr | USA | On-site – No longer accepting applications As an Information Security – Firewall & Network Engineer, you will be responsible for the design, implementation, maintenance, and support of the organization’s firewall infrastructure, primarily focusing on Palo Alto Networks technologies. The engineer will ensure the security and integrity of the company’s networks and will work closely with the IT security team to develop comprehensive network security measures. Information Security Officer Dubai Civil Aviation Authority | UAE | On-site – No longer accepting applications As a Information Security Officer, you will identify vulnerabilities in DCAA’s network, develop and implement security plans, and monitor network usage for policy compliance. You’ll stay current with IT security threats and standards and perform penetration tests to detect flaws. Information Security Penetration Tester Pennymac | USA | On-site – No longer accepting applications As an Information Security Penetration Tester, you will conduct security assessments using a combination of manual and automated techniques to uncover vulnerabilities and potential attack vectors across enterprise assets. Identify and prioritize critical vulnerabilities, incorporating threat intelligence to assess potential exploitability and ensure remediation efforts align with the evolving threat landscape. Information Technology and Cyber Risk Manager HSBC | Germany | Hybrid – No longer accepting applications As an Information Technology and Cyber Risk Manager, you will define security policies, support penetration testing, and drive regulatory and strategic programs. You will also provide technical support for projects such as new core banking systems and the DORA transition. Junior Cybersecurity Operations Center Analyst Philippine Airlines | Philippines | Hybrid – No longer accepting applications As a Junior Cybersecurity Operations Center Analyst, you will monitor network traffic, system logs, and security devices in a 24×7 SOC environment, analyzing anomalous activity and investigating security incidents. Evaluate security alerts, prioritize incidents, and escalate critical issues, providing incident response support and following standard procedures to mitigate security risks. Mainframe Security Lead AXA Group Operations | France | On-site – View job details As a Mainframe Security Lead, you will assess IBM Mainframe Z Series and iSeries environments for vulnerabilities and coordinate incident response with relevant teams. You will identify potential threats and take proactive measures to mitigate risks. Penetration Testing Analyst Sun Life | Ireland | On-site – No longer accepting applications As a Penetration Testing Analyst, you will perform penetration testing and vulnerability scanning activities for Sun Life’s internal and external clients. Present reports and data on the findings from penetration testing to the various support teams responsible for these applications. Provide guidance for these teams, making recommendations and/or delivering recommendations to ensure any vulnerabilities found are adequately addressed. Platform Security Engineer – PKI & Cryptography Generali | Italy | Hybrid – No longer accepting applications As a Platform Security Engineer – PKI & Cryptography, you will design, deploy, and manage PKI services, including root and subordinate Certificate Authorities. Oversee the lifecycle of digital certificates managing the CLM solution of the Group. Administer and maintain HSMs, ensuring secure key storage and operations. Support the management of cryptographic solutions (symmetric/asymmetric encryption, key management, signing, hashing). Senior Application Security Engineer Kaltura | Israel | Hybrid – No longer accepting applications As a Senior Application Security Engineer, you will work with product and engineering teams to embed security across the SDLC. You will conduct design reviews, threat modeling, code reviews, and drive remediation while managing security tools like SAST, SCA, and IaC. You will lead penetration tests, mentor developers on best practices, and implement automation frameworks and product security features. You will also support RFPs and customer security requests. Senior IAM Specialist Sobeys | Canada | Hybrid – No longer accepting applications As a Senior IAM Specialist, you will design, deploy and manage CyberArk Secure Infrastructure Access (SIA) for secure, credential-less access to Linux, Windows and Databases across hybrid environments. Implement and maintain Identity Security Intelligence (ISI) for behavioral analytics, anomaly detection, and risk-based policy enforcement. Configure and manage CyberArk Remote Access for secure, agentless vendor access without VPN dependency. Senior Red Team Specialist CyberGate Defense | UAE | On-site – View job details As a Senior Red Team Specialist, you will design and lead red team engagements that simulate real-world threat actors. Perform comprehensive internal and external penetration tests across web, mobile, and network infrastructure. Identify vulnerabilities, develop custom exploits, and escalate privileges.Conduct social engineering assessments, including phishing, vishing, and physical intrusion tests. Senior Security Architect Keyrock | France | Remote – No longer accepting applications As a Senior Security Architect, you will design security methods and controls, ensuring/enforcing sensitive data/asset and system security, integrity, and confidentiality. Identify and assess security risks and vulnerabilities, and formulate/implement mitigation/enforcement strategies. Develop, refine, and maintain security policies, standards, and procedures emphasizing data and funds security. SOC Operator Crisis24 | USA | On-site – No longer accepting applications As a SOC Operator, you will monitor and assess daily alerts, reported incidents, and intrusion alarms. You’ll manage access control, video systems, and intrusion detection, dispatching security officers as needed. You’ll also monitor events using open-source intelligence and coordinate with local assets. Vulnerability Remediation Engineer ECI | Philippines | On-site – No longer accepting applications As a Vulnerability Remediation Engineer, you will lead complex remediation efforts across infrastructure, applications, and cloud environments. Analyze vulnerability scan results, assess risk, and recommend or implement appropriate remediation actions. Ensure remediation activities comply with internal policies, SLAs, and regulatory requirements. Vulnerability Management Specialist Grant Thornton | Ireland | On-site – No longer accepting applications As a Vulnerability Management Specialist, you will manage and maintain the enterprise vulnerability management program, ensuring timely identification and remediation of vulnerabilities across on-prem, cloud, and hybrid environments. Perform vulnerability scans, secure configuration scans, and policy compliance scans using Qualys and other tools.

Automotive manufacturing giant Stellantis has confirmed that attackers stole some of its North American customers' data after gaining access to a third-party service provider's platform. Stellantis is a multinational corporation formed in 2021 after the merger of the PSA Group (Peugeot Société Anonyme) and Fiat Chrysler Automobiles (FCA). Stellantis is currently one of the largest automotive companies globally by revenue and the world's fifth-largest automaker by volume. The company owns 14 major automotive brands, including Alfa Romeo, Chrysler, Citroën, Dodge, DS Automobiles, Fiat, Jeep, Lancia, Maserati, Opel, Peugeot, Ram, and Vauxhall, and it operates manufacturing facilities across Europe, North America, South America, and other regions, with operations in over 130 countries. According to a statement published over the weekend, the attackers only stole customer contact information during the breach since the compromised platform was not used to store financial or other sensitive personal information. "We recently detected unauthorized access to a third-party service provider's platform that supports our North American customer service operations," Stellantis said. "Upon discovery, we immediately activated our incident response protocols, initiated a comprehensive investigation, and took prompt action to contain and mitigate the situation. We are also notifying the appropriate authorities and directly informing affected customers." The auto giant also advised customers to be cautious of potential phishing attempts and to refrain from clicking suspicious links or sharing personal information when receiving unexpected emails, texts, or calls. BleepingComputer reached out to Stellantis with questions about the incident, but a response was not immediately available. Salesforce data breach claimed by ShinyHunters Although Stellantis didn't share more information regarding this attack, BleepingComputer has learned that it is part of a recent wave of Salesforce data breaches linked with the ShinyHunters extortion group, which has affected numerous high-profile companies. Earlier today, ShinyHunters claimed responsibility for the Stellantis data breach and told BleepingComputer that they had stolen over 18 million Salesforce records, including names and contact details, from the company's Salesforce instance. Since the start of the year, the extortion group has been targeting Salesforce customers in data theft attacks using voice phishing attacks, impacting companies such as Google, Cisco, Qantas, Adidas, Allianz Life, Farmers Insurance, Workday, and LVMH subsidiaries, including Dior, Louis Vuitton, and Tiffany & Co. ShinyHunters also claims they used stolen OAuth tokens for Salesloft's Drift AI chat integration with Salesforce to steal sensitive information, such as passwords, AWS access keys, and Snowflake tokens, after gaining access to customers' Salesforce instances. Using this method, they claimed to have stolen customer information from Google, Cloudflare, Zscaler, Tenable, Palo Alto Networks, CyberArk, Nutanix, Qualys, Rubrik, Elastic, BeyondTrust, Proofpoint, JFrog, Cato Networks, and many more. Last week, the FBI released a Flash alert sharing IOCs discovered during the attacks and warning about threat actors breaching organizations' Salesforce environments to steal data and extort victims. Meanwhile, the extortion group told BleepingComputer that they had stolen over 1.5 billion Salesforce records from 760 companies, using compromised Salesloft Drift OAuth tokens. Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply. The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments. Get the report

The ShinyHunters extortion group claims to have stolen over 1.5 billion Salesforce records from 760 companies using compromised Salesloft Drift OAuth tokens. For the past year, the threat actors have been targeting Salesforce customers in data theft attacks using social engineering and malicious OAuth applications to breach Salesforce instances and download data. The stolen data is then used to extort companies into paying a ransom to prevent the data from being publicly leaked. These attacks have been claimed by threat actors stating they are part of the ShinyHunters, Scattered Spider, and Lapsus$ extortion groups, now calling themselves "Scattered Lapsus$ Hunters." Google tracks this activity as UNC6040 and UNC6395. In March, one of the threat actors breached Salesloft's GitHub repository, which contained the private source code for the company. ShinyHunters told BleepingComputer that the threat actors used the TruffleHog security tool to scan the source code for secrets, which resulted in the finding of OAuth tokens for the Salesloft Drift and the Drift Email platforms. Salesloft Drift is a third-party platform that connects the Drift AI chat agent with a Salesforce instance, allowing organizations to sync conversations, leads, and support cases into their CRM. Drift Email is used to manage email replies and organize CRM and marketing automation databases. Using these stolen Drift OAuth tokens, ShinyHunters told BleepingComputer that the threat actors stole approximately 1.5 billion data records for 760 companies from the "Account", "Contact", "Case", "Opportunity", and "User" Salesforce object tables. Of these records, approximately 250 million were from the Account, 579 million from Contact, 171 million from Opportunity, 60 million from User, and about 459 million records from the Case Salesforce tables. The Case table was used to store information and text from support tickets submitted by customers of these companies, which, for tech companies, could include sensitive data. As proof that they were behind the attack, the threat actor shared a text file listing the source code folders in the breached Salesloft GitHub repository. BleepingComputer contacted Salesloft with questions about these record counts and the total number of companies impacted, but did not receive a response to our email. However, a source confirmed that the numbers are accurate. Google Threat Intelligence (Mandiant) reported that the stolen Case data was analyzed for hidden secrets, such as credentials, authentication tokens, and access keys, to enable the attackers to pivot into other environments for further attacks. "After the data was exfiltrated, the actor searched through the data to look for secrets that could be potentially used to compromise victim environments," explained Google. "GTIG observed UNC6395 targeting sensitive credentials such as Amazon Web Services (AWS) access keys (AKIA), passwords, and Snowflake-related access tokens." The stolen Drift and Drift Email tokens were used in large-scale data theft campaigns that hit major companies, including Google, Cloudflare, Zscaler, Tenable, CyberArk, Elastic, BeyondTrust, Proofpoint, JFrog, Nutanix, Qualys, Rubrik, Cato Networks, Palo Alto Networks, and many more. Due to the sheer volume of these attacks, the FBI recently released an advisory warning about the UNC6040 and UNC6395 threat actors, sharing IOCs discovered during the attacks. Last Thursday, the threat actors claiming to be part of Scattered Spider stated that they planned to "go dark" and stop discussing operations on Telegram. In a parting post, the threat actors claimed to have breached Google's Law Enforcement Request system (LERS), which is used by law enforcement to issue data requests, and the FBI eCheck platform, used for conducting background checks. After contacting Google about these claims, the company confirmed that a fraudulent account was added to its LERS platform. "We have identified that a fraudulent account was created in our system for law enforcement requests and have disabled the account," Google told BleepingComputer. "No requests were made with this fraudulent account, and no data was accessed." While the threat actors indicated they are retiring, researchers from ReliaQuest report that the threat actors began targeting financial institutions in July 2025 and are likely to continue conducting attacks. To protect against these data theft attacks, Salesforce recommends that customers follow security best practices, including enabling multi-factor authentication (MFA), enforcing the principle of least privilege, and carefully managing connected applications. Overall prevention scores can hide what happens after initial access. Once attackers are using valid credentials, prevention drops sharply. The Blue Report 2026 measures defenses technique by technique across 338 million simulations run in customer production environments. Get the report

Threat actors had access to Salesloft’s GitHub account between March and June 2025 and performed reconnaissance in preparation for the widespread Salesforce-Salesloft data theft campaign. The data breach occurred between August 8 and August 18, when the attackers used compromised OAuth tokens for the Drift AI chatbot to export large volumes of data from Salesforce environments. Attributed to a threat actor tracked as UNC6395, the campaign hit hundreds of organizations and focused on the extraction of AWS access keys, passwords, and Snowflake-related access tokens from the stolen data. Initially believed to affect only accounts using the Salesforce-Salesloft Drift integration, the attack was later found to have affected other entities as well, including Google Workspace customers. The attack resulted in Salesforce disabling the Salesloft integration, and in Drift being taken temporarily offline to improve its security. On September 7, the Salesforce-Salesloft integration was restored. However, the campaign was not the result of a weakness in Drift, Salesloft said on Sunday. Instead, it was possible because hackers had compromised the company’s GitHub account half a year ago. “In March through June 2025, the threat actor accessed the Salesloft GitHub account. With this access, the threat actor was able to download content from multiple repositories, add a guest user and establish workflows,” Salesloft revealed. The investigation into the incident, performed by Mandiant, revealed that the hackers performed reconnaissance in the Salesloft and Drift application environments, and then accessed Drift’s AWS instance, exfiltrating OAuth tokens for customers’ integrations. “The threat actor used the stolen OAuth tokens to access data via Drift integrations,” Salesloft says. According to the company, the attack has been contained and the attackers evicted from its environments, and Mandiant has validated that. What Salesloft did not specify, however, was the number of impacted organizations. According to previous estimations, roughly 700 companies might have been affected. In the cybersecurity space, Cloudflare, Palo Alto Networks, and Zscaler were the first to confirm impact from the attack, followed shortly by Proofpoint, SpyCloud, Tanium, and Tenable. The list of cybersecurity firms impacted by the incident, however, has grown to over a dozen, and also includes BeyondTrust, Bugcrowd, CyberArk, Cato Networks, JFrog, PagerDuty, and Rubrik. Elastic said a single email account was compromised through the ‘Drift Email’ integration. Esker, Heap, Megaport, Nutanix, Sigma Computing, and Workiva were also hit, Nudge Security reveals. In most cases, the compromised Salesforce instances stored data related to customer support tickets, including business information such as names, email addresses, and phone numbers. Related: Over 6,700 Private Repositories Made Public in Nx Supply Chain Attack Related: How to Close the AI Governance Gap in Software Development Related: PLoB: A Behavioral Fingerprinting Framework to Hunt for Malicious Logins

Cybersecurity researchers have discovered over a dozen vulnerabilities in enterprise secure vaults from CyberArk and HashiCorp that, if successfully exploited, can allow remote attackers to crack open corporate identity systems and extract enterprise secrets and tokens from them. The 14 vulnerabilities, collectively named Vault Fault, affect CyberArk Secrets Manager, Self-Hosted, and Conjur Open Source and HashiCorp Vault, according to a report from an identity security firm Cyata. Following responsible disclosure in May 2025, the flaws have been addressed in the following versions - CyberArk Secrets Manager and Self-Hosted 13.5.1 and 13.6.1 CyberArk Conjur Open Source 1.22.1 HashiCorp Vault Community Edition 1.20.2 or Vault Enterprise 1.20.2, 1.19.8, 1.18.13, and 1.16.24 These include authentication bypasses, impersonation, privilege escalation bugs, code execution pathways, and root token theft. The most severe of the issues allows for remote code execution, allowing attackers to takeover the vault under certain conditions without any valid credentials - CVE-2025-49827 (CVSS score: 9.1) - Bypass of IAM authenticator in CyberArk Secrets Manager CVE-2025-49831 (CVSS score: 9.1) - Bypass of IAM authenticator in CyberArk Secrets Manager via a misconfigured network device CVE-2025-49828 (CVSS score: 8.6) - Remote code execution in CyberArk Secrets Manager CVE-2025-6000 (CVSS score: 9.1) - Arbitrary remote code execution via plugin catalog abuse in HashiCorp Vault CVE-2025-5999 (CVSS score: 7.2) - Privilege escalation to root via policy normalization in HashiCorp Vault In addition, vulnerabilities have also been discovered in HashiCorp Vault's lockout protection logic, which is designed to throttle brute-force attempts, that could permit an attacker to infer which usernames are valid by taking advantage of a timing-based side channel and even reset the lockout counter by changing the case of a known username (e.g., admin to Admin). Two other shortcomings identified by the Israeli company made it possible to weaken lockout enforcement and bypass multi-factor authentication (MFA) controls when username_as_alias=true in the LDAP auth configuration and MFA enforcement is applied at the EntityID or IdentityGroup level. In the attack chain detailed by the cybersecurity company, it's possible to leverage a certificate entity impersonation issue (CVE-2025-6037) with CVE-2025-5999 and CVE-2025-6000 to break the authentication layer, escalate privileges, and achieve code execution. CVE-2025-6037 and CVE-2025-6000 are said to have existed for over eight and nine years, respectively. Armed with this capability, a threat actor could further weaponize the access to delete the "core/hsm/_barrier-unseal-keys" file, effectively turning a security feature into a ransomware vector. What's more, the Control Group feature can be undermined to send HTTP requests and receive responses without being audited, creating a stealthy communication channel. "This research shows how authentication, policy enforcement, and plugin execution can all be subverted through logic bugs, without touching memory, triggering crashes, or breaking cryptography," security researcher Yarden Porat said. In a similar vein, the vulnerabilities discovered in CyberArk Secrets Manager/Conjur allow for authentication bypass, privilege escalation, information disclosure, and arbitrary code execution, effectively opening the door to a scenario where an attacker can craft an exploit chain to obtain unauthenticated access and run arbitrary commands. The attack sequence unfolds as follows - IAM authentication bypass by forging valid-looking GetCallerIdentity responses Authenticate as a policy resource Abuse the Host Factory endpoint to create a new host that impersonates a valid policy template Assigned a malicious Embedded Ruby (ERB) payload directly to the host Trigger the execution of the attached ERB by invoking the Policy Factory endpoint "This exploit chain moved from unauthenticated access to full remote code execution without ever supplying a password, token, or AWS credentials," Porat noted.

Le minacce informatiche sono sempre più sofisticate, frequenti e dannose. Secondo Statista, entro il 2029, il costo globale dei danni provocati dalla criminalità informatica raggiungerà i 15,6 trilioni di dollari, esercitando una crescente pressione sulle organizzazioni affinché rafforzino le proprie difese. Di conseguenza, le aziende sono chiamate ad aumentare in modo significativo gli investimenti nella cyber security. Gartner stima che la spesa globale per la sicurezza informatica raggiungerà i 212 miliardi di dollari nel 2025, registrando un incremento del 15,1% rispetto al 2024. Inoltre, l’adozione diffusa di dispositivi IoT (Internet of Things) e l’implementazione delle reti 5G e 6G ampliano la superficie di attacco, oltre a creare nuove vulnerabilità che richiedono strategie avanzate per proteggere efficacemente i sistemi interconnessi. Ancora, la carenza di professionisti qualificati in sicurezza informatica costituisce un’ulteriore problema critico da gestire, ostacolando la capacità delle organizzazioni di implementare e di mantenere misure di sicurezza efficaci, esponendole a potenziali rischi cyber. Indice degli argomenti La sicurezza informatica scaturisce dall’insieme delle misure tecniche, organizzative e procedurali volte a proteggere sistemi informativi, dati, reti e identità digitali – sia individuali sia aziendali – da attacchi cyber, oppure a mitigarne l’impatto. Tale protezione si fonda sull’adozione di tecnologie di sicurezza informatica all’avanguardia, sull’implementazione di politiche di protezione rigorose, sul mantenimento di un elevato livello di consapevolezza all’interno dell’organizzazione, oltre ad un aggiornamento costante sulle minacce emergenti e sulle migliori pratiche di difesa. I software di sicurezza rappresentano strumenti fondamentali per la protezione informatica ed esistono sul mercato diverse tipologie di soluzioni in grado di: monitorare costantemente l’infrastruttura IT, individuare potenziali minacce e mitigarle; supportare l’identificazione e la risoluzione tempestiva delle vulnerabilità, prima che possano compromettere i sistemi. Di seguito un elenco delle principali tipologie di software di sicurezza e loro funzionalità: software di monitoraggio della sicurezza della rete per monitorare e bloccare le minacce; scanner di vulnerabilità per individuare le vulnerabilità nei tuoi sistemi ed eliminarle in tempo; firewall per filtrare il traffico dannoso e impedirne l’accesso alla rete; strumenti di crittografia per proteggere le comunicazioni e i contenuti; software antivirus per impedire a virus informatici e worm di infettare i tuoi dispositivi; sistemi di Intrusion detection system/intrusion prevention system (IDS/IPS) per rilevare e prevenire le minacce; strumenti di penetration testing per valutare l’efficacia delle misure di sicurezza. Di fatto, questi strumenti di cybersecurity garantiscono all’organizzazione diversi vantaggi in termini di: protezione dei dati personali; monitoraggio continuo e sicurezza; protezione del brand e mantenimento della fiducia di clienti, di partner e di investitori; miglioramento della produttività dei dipendenti; gestione dei requisiti normativi e di conformità, attraverso funzioni di analisi e di audit; miglioramento della sicurezza della forza lavoro distribuita o remota; gestione delle identità e delle autorizzazioni di accesso su sistemi e dati critici. risparmio in termini di costi di recovery. È importante selezionare il giusto fornitore di software di sicurezza informatica che comprenda l’azienda in termini di: specifico profilo di rischio, obiettivi aziendali e vincoli operativi. Di seguito una mini roadmap che fornisce aspetti fondamentali da considerare per una scelta ponderata e consapevole di una soluzione di sicurezza informatica: definire il proprio livello di sicurezza; valutare le capacità tecniche dei fornitori; valutare la scalabilità e l’integrazione; valutazione del servizio di supporto e la competenza; considerare il costo totale del software di sicurezza. È fondamentale iniziare valutando il proprio contesto di sicurezza, considerando se si è in cloud-native, ibrido o ancora prevalentemente on-premise. Oppure se si hanno team distribuiti, requisiti di conformità o normative specifiche di settore a cui conformarsi. Ovvero si tratta di considerare questi fattori essenziali: architettura dell’infrastruttura: le infrastrutture cloud-native, per esempio, richiedono modelli di sicurezza dinamici, scalabili e basati su API, a differenza dei tradizionali sistemi on-premise, che spesso si basano su protezioni statiche basate sul perimetro e livelli hardware fissi; requisiti di conformità: settori come la finanza, la sanità e la pubblica amministrazione devono rispettare normative rigorose, rendendo essenziali strumenti di protezione incentrati sulla conformità per evitare sanzioni costose; capacità del team: il personale interno deve possedere competenze in materia di sicurezza informatica per identificare le lacune in cui sono necessari applicazioni di sicurezza, o competenze esterne, per il monitoraggio 24 ore su 24, 7 giorni su 7, la risposta agli incidenti o il supporto specializzato per la conformità. I principali player di sicurezza informatica oggi combinano diversi aspetti di cybersecurity in piattaforme unificate. A tal proposito si consiglia di verificare la disponibilità di strumenti di sicurezza, quali: intelligenza artificiale comportamentale e apprendimento automatico: le minacce moderne richiedono sistemi in grado di apprendere e adattarsi. Pertanto, si consiglia di valutare fornitori di software di sicurezza informatica che utilizzino l’intelligenza artificiale per rilevare anomalie e prevedere potenziali attacchi, anziché limitarsi a rispondere a minacce note; rilevamento e risposta estesi (XDR – Extended Detection and Response): la soluzione di sicurezza dovrebbe fornire una visibilità unificata su endpoint, reti, carichi di lavoro cloud e identità degli utenti, dato che strumenti di sicurezza frammentati creano “zone d’ombra” che gli aggressori possono sfruttare; Zero Trust: si tratta un’architettura in grado di verificare ogni richiesta di accesso in base al contesto ed in tempo reale, inclusi il comportamento dell’utente, lo stato del dispositivo e la posizione, dato che la tradizionale sicurezza perimetrale è obsoleta negli ambienti distribuiti odierni. Si tratta di verificare la capacità della soluzione di sicurezza informatica di adeguarsi, nel tempo, alle esigenze dell’organizzazione, senza richiedere la sostituzione completa del software o della piattaforma. Pertanto, si consiglia di verificare i seguenti aspetti: integrazione cloud nativa: optare per strumenti di sicurezza che si integrano direttamente con i provider e servizi cloud, consentendo visibilità in tempo reale, controlli automatizzati e protezione coerente in ambienti multi-cloud e ibridi; architettura API-First: scegliere soluzioni progettate con API aperte, che consentano un’integrazione perfetta con gli strumenti, i flussi di lavoro e le applicazioni personalizzate esistenti per mantenere la continuità operativa e l’automazione; automazione delle policy: optare per soluzioni che applichino e facciano rispettare le policy di sicurezza in modo dinamico in tutti gli ambienti, garantendo la conformità, oltre a ridurre l’intervento manuale man mano che l’infrastruttura si espande o cambia rapidamente. I I fornitori di soluzioni di sicurezza informatica devono essere i grado di fornire sia tecnologia sia competenze, pertanto, si consiglia di verificare la disponibilità di servizi quali: supporto e monitoraggio 24 ore su 24, 7 giorni su 7: il fornitore deve garantire capacità di rilevamento delle minacce in tempo reale e di risposta rapida 24 ore su 24, compresi i giorni festivi e gli orari non lavorativi; pianificazione della risposta agli incidenti: un valido provider di soluzioni di sicurezza è in grado di supportare l’organizzazione attraverso piani di risposta agli incidenti ben documentati, garantendo un rapido contenimento, ripristino e comunicazione durante le violazioni; formazione continua: i principali player offrono formazione continua sulla sicurezza e aggiornamenti sulle minacce, consentendo all’organizzazione di riconoscere nuovi vettori di attacco, oltre a rafforzare le best practice tramite moduli di apprendimento pratici basati sui ruoli. Sebbene la sicurezza sia fondamentale, deve essere sostenibile. Pertanto, l’organizzazione deve valutare attentamente: costi di licenza e abbonamento: valutare se la struttura dei prezzi della soluzione di sicurezza informatica si adatta in modo efficiente alla crescita dell’organizzazione in termini di: numero di utenti; livelli di funzionalità; potenziali implicazioni sui costi a lungo termine per aggiornamenti o per rinnovi; implementazione e formazione: considerare l’impegno tecnico, il personale interno e il tempo necessari per implementare completamente la soluzione, oltre a formare gli utenti e integrarla senza problemi nei sistemi e nei flussi di lavoro esistenti; ratio costo/opportunità: considerare i rischi finanziari, reputazionali e operativi associati a potenziali violazioni della sicurezza rispetto al valore dell’investimento. In base alle recensioni di esperti di settore disponibili online, ecco un elenco dei principali fornitori di soluzioni di sicurezza informatica, noti per la loro popolarità e l’ampiezza delle funzionalità offerte (in ordine alfabetico): Akamai Armis Checkpoint Cisco Cloudfire Crowdstrike CyberArk Darktrace Fortinet Palo Alto Networks Proofpoint SentinelOne ServiceNow Symantec Sophos Splunk Sprinto Trend Micro.

Identity security and access management firm CyberArk has patched several serious vulnerabilities that could lead to unauthenticated remote code execution, potentially enabling threat actors to gain access to valuable enterprise secrets. The vulnerabilities were found by researchers at agentic identity security firm Cyata in CyberArk Conjur, an open source secrets management solution that is used by many organizations for managing machine and AI identities and for brokering secure access between various enterprise environments. Conjur is designed for securely storing, managing and controlling access to credentials, certificates, API keys and other enterprise secrets used in cloud and DevOps environments, which could be highly valuable to threat actors. Cyata discovered a series of vulnerabilities, including ones allowing IAM authentication bypass, privilege escalation, information disclosure, and arbitrary code execution. Chaining the flaws enables a remote, unauthenticated attacker to execute arbitrary code on the targeted system without needing any password, token or AWS credentials. [ Read: Palo Alto Networks to Acquire CyberArk for $25 Billion ] The vulnerabilities are tracked as CVE-2025-49827, CVE-2025-49831 (both IAM authenticator bypasses), CVE-2025-49828 (remote code execution), CVE-2025-49830 (path traversal and file disclosure), and CVE-2025-49829 (missing validations). CyberArk was notified about the findings in late May and the company announced the availability of patches in a blog post published on July 15. Customers had previously been notified about the flaws and patches. CyberArk’s Secrets Manager, Self-Hosted (formerly Conjur Enterprise) and Conjur open source are affected. “As far as we know, these vulnerabilities have not been exploited in the wild, but we strongly encourage all users of the affected software to deploy the newly released patches as soon as possible,” CyberArk said. In addition to the CyberArk product vulnerabilities, Cyata researchers discovered flaws in another widely used secrets management platform, HashiCorp Vault. A total of nine vulnerabilities were found, some of them allowing remote code execution and a full system takeover. Cyata presented its findings on Wednesday at the Black Hat conference and disclosed technical details in a blog post. Related: Flaws Expose 100 Dell Laptop Models to Implants, Windows Login Bypass Related: Reclaiming Control: How Enterprises Can Fix Broken Security Operations Related: Flaw in Vibe Coding Platform Base44 Exposed Private Enterprise Applications

Many companies are showcasing their cybersecurity products and services this week at the 2025 edition of the Black Hat conference in Las Vegas. To help cut through the clutter, the SecurityWeek team is publishing a digest summarizing some of the announcements made by vendors at Black Hat USA 2025, including new products and services, updates to existing offerings, reports, and other initiatives. This is part two of the roundup. The first part covers announcements made in the days leading up to the event, as well as some of the announcements made on Monday, August 4. 1Password study 1Password has announced new findings from a survey of North American security leaders on AI usage and emerging threats. Nearly two-thirds (63%) of security leaders feel the biggest internal security threat is their employees unknowingly giving AI agents access to sensitive data. Additionally, 50% say their organizations have experienced a confirmed or suspected cyber incident caused by AI or AI agents in the last six months. Arctic Wolf announces new integrations Arctic Wolf has announced new integrations with Microsoft, Oracle, OneLogin, and CyberArk. These new integrations expand the platform’s openness and flexibility, enabling organizations to address the challenges of fragmented security tools and increasingly dynamic environments. AttackIQ launches Watchtower Adversarial exposure validation firm AttackIQ has launched Watchtower, an AI-powered threat intelligence analyzer designed to provide visibility into the threat actors actively targeting enterprises. AttackIQ Watchtower hyperlocalizes threat intelligence and generates tailored emulation scenarios to deliver contextual insights. Black Duck unveils enhancements to application security assistant Black Duck Software has announced enhancements to Duck Assist, its AI-powered application security assistant. The new capabilities bring automated security scanning for AI-generated code and AI-driven code fixes directly into developer environments. Cyware launches open source AI integration server Cyware unveiled a major expansion to its AI-powered threat intelligence ecosystem with the launch of Cyware Model Context Protocol (MCP) Server, an open source capability that enables security teams to connect LLMs and AI assistants directly into their threat intel and automation workflows. The new MCP Server supports natural language-driven actions and investigations using secure AI agents, seamless integration with customer-owned LLMs, and full context and control across detection, enrichment, and response. Darktrace cyber threat landscape report Darktrace released a retrospective of the first six months of 2025, highlighting key trends across the threat landscape. Darktrace observed a wide range of tactics during this period, used by various types of threat actors, including APTs, Malware-as-a-Service (MaaS) and Ransomware-as-a-Service (RaaS) groups. Flashpoint releases new AI search and investigation capabilities Threat intelligence firm Flashpoint has released two new AI-powered capabilities for its flagship platform Flashpoint Ignite: AI Summarization for Search and AI Summarization for Investigations. AI Summarization for Search distills thousands of posts into clear, actionable insight in seconds. AI Summarization for Investigations is an automation that transforms investigations from an evidence-collection point into a dynamic, insight-generating engine. Menlo launches Secure Storage Menlo Security launched Menlo Secure Storage and Menlo Adaptive Web Modules. These two new offerings are designed to give organizations greater control over user experiences and sensitive data. Menlo Secure Storage allows users to upload and download files between trusted applications without saving anything locally. Menlo Adaptive Web Modules is a suite of cloud-based browser controls that gives organizations fine-grained control over how users interact with web content. NetRise platform update Software supply chain security firm NetRise has introduced a significant update to its core product platform to help users prioritize, mitigate, and remediate vulnerabilities found in the software they develop. New features include context on whether a vulnerability is reachable, SBOM editing, fix version data, and platform re-architecture. Netskope expands platform Network security company Netskope has unveiled two platform advancements. Netskope One Copilot for Private Access is an AI-powered assistant that optimizes universal zero trust network access deployment. Netskope Model Context Protocol (MCP) server, currently in preview, provides a secure bridge that enables LLMs like Microsoft Copilot, Claude Desktop, Google Vertex and Amazon Bedrock to interact directly and safely with Netskope APIs, enabling LLMs to gain context from a customer’s Netskope environment. PointGuard AI expands platform AI security and governance firm PointGuard AI announced an expansion of its platform to provide AI discovery, threat correlation, and protection across the full AI stack. The platform now scans GitHub and other source code repositories to identify AI-related components including models, datasets, notebooks, API calls, and libraries, along with connections to external applications or data sources. Reveal Security unveils platform Identity security company Reveal Security has launched the Reveal Platform, which provides complete visibility into identity actions, both human and non-human, across all SaaS, cloud, and custom applications. This complete view gives security teams the ability to preemptively take action before a security incident or breach occurs. SandboxAQ launches AQtive Guard Protect SandboxAQ announced AQtive Guard Protect, a solution designed to secure the full lifecycle of non-human identities and cryptographic assets. The solution provides deep observability into machine identities across multi-cloud, hybrid, and containerized environments. It also provides automation to reduce the overhead of managing and rotating secrets, as well as last-mile governance to ensure high-risk credentials don’t slip through the cracks. Semperis offering protects highly targeted non-human identities Semperis launched Service Account Protection Essential, a new edition of its Directory Services Protector, designed to tackle one of the most persistent threats in Active Directory environments: service accounts. The new tool helps organizations identify service accounts, build an inventory, and continually monitor them for security vulnerabilities. It also discovers unknown and misplaced service accounts, identifies risky configurations and critical exposures, detects stale and misconfigured accounts, and alerts on malicious and anomalous behavior. SOCRadar launches agentic threat intelligence platform SOCRadar launched SOCRadar Agentic Threat Intelligence, a new platform that automates threat intelligence through the deployment of autonomous AI agents that proactively detect, analyze, and respond to external threats with minimal human intervention. XM Cyber integration with Google Cloud XM Cyber announced that its Continuous Exposure Management (CEM) platform is now fully embedded into Google Security Operations, including its cloud-native SIEM and SOAR. This gives enterprises real-world exposure data to enrich threat detection—helping SOC teams cut through noise and act faster. Xona platform update Secure access solutions provider Xona announced the general availability of Xona Platform v5.4.2, a major update that extends centralized control, scalable auditability, and unified policy enforcement across globally distributed operational environments. ZEST Security adds AWS Service Control Policies ZEST Security announced the addition of AWS Service Control Policies (SCPs) as part of its broader exposure mitigation offering. This gives security teams a new, code-free way to proactively block attacker activity and reduce cloud exposure. By adding SCPs into its mitigation offering, ZEST is allowing security teams to proactively limit attacker activity and reduce exposure without waiting on patches and code changes.

AI and ml Payments giant Stripe is about to drop over $7 billion to become a gateway to AI token salesAI gateways look promising as companies struggle with model orchestration ai and ml Anthropic says text watermarking scheme relies on inconsequential words'Shall I compare thee to a summer's afternoon' is the sort of thing this will make, and others look likely to adopt it AI and ML DeepSeek's innovative harness treats everything as a plug-inChinese AI labs keep moving forward while US labs play defense SECURITY Autonomous AI attacks pose 'clear and present danger' to critical infrastructureWeaponized agents could turn digital intrusions into kinetic disasters, experts warn off-prem Rent-a-GPU outfit Nebius promises rapid 1 GW powerup plan isn't nebulousAnd it'll jump through every financial hoop it can to get there Security Russians are posing as Signal support to launch phishing attacksPLUS: US takes down Iranian propaganda sites; Marketing company asks 'Why Do We Have Your Information?' And more! Security Microsoft patches failed to fix on-prem SharePoint, which is now under zero-day attackPLUS: China upgrades smartphone surveillance tools; Ring eases anti-snooping stance; and more Black Hat and DEF CON DEF CON Franklin project enlists hackers to harden critical infrastructureVoting village reports have been so successful, says Jeff Moss, that the whole of DEF CON will now be included Security EQT buys majority share in Swiss cybersecurity biz AcronisWent at equivalent of $3.5B+ valuation for entire firm, though portion sold not specified Malware Month Ten years since the first corp ransomware, Mikko Hyppönen sees no end in sightOn the plus side, infosec's a good bet for a long, stable career FOSS smashed one Microsoft monopoly. After 20 years of failure, it's time to smash anotherWord up GNOME can look like Windows – and Flashback can do it without extensionsNew 'Simple-taskbar' is an option, but there's a simpler, stabler way A moment of silence, please, for the final release of Debian on x86-32New Debian versions hit FOSSland in the form of 13.6 and 12.15 Baddies caught exploiting extensions bugs with perfect 10 scores on vulnerable Joomla websitesFlaws in iCagenda, Balbooa Forms extensions can impact open source CMS that powers a million sites worldwide Frame: A new X11 server – implemented directly in assemblyJoins yserver, Phoenix, and of course XLibre – and outlier Arcan Cinnamon 6.8 will support Wayland – if you want itNext version of Linux Mint’s desktop has both kinds of display server

Palo Alto Networks to acquire identity security provider CyberArk in $25 billion deal Cybersecurity giant Palo Alto Networks will acquire the Israeli identity security firm CyberArk in a deal worth about $25 billion, the companies announced on Wednesday. Under the terms of the agreement, Palo Alto Networks will pay investors of the publicly listed CyberArk $45 per share, as well as more than two shares of the Silicon Valley company’s stock. CyberArk shares jumped 13% on Tuesday after Wall Street Journal reported a deal was close, while shares in Palo Alto Networks have fallen more than 10%. Palo Alto CEO Nikesh Arora said in a statement that the acquisition is an attempt to get ahead of demand for identity security tools, which manage access to systems, in the artificial intelligence age. "Our market entry strategy has always been to enter categories at their inflection point, and we believe that moment for Identity Security is now,” he said. CyberArk’s tools will be used to manage permissions for AI agents as well as people, the companies said in a press release. “Securing privileged credentials for both human and machine identities—including AI agents and workloads—is the most critical challenge for the modern enterprise,” they said. The deal is one of the year’s biggest in the cybersecurity business, second only to Google’s agreement in March to acquire the cloud security startup Wiz for $32 billion. That deal is under regulatory review and not expected to close until 2026. James Reddick has worked as a journalist around the world, including in Lebanon and in Cambodia, where he was Deputy Managing Editor of The Phnom Penh Post. He is also a radio and podcast producer for outlets like Snap Judgment.

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) added two security flaws impacting SysAid IT support software to its Known Exploited Vulnerabilities (KEV) catalog, based on evidence of active exploitation. The vulnerabilities in question are listed below - CVE-2025-2775 (CVSS score: 9.3) - An improper restriction of XML external entity (XXE) reference vulnerability in the Checkin processing functionality, allowing for administrator account takeover and file read primitives CVE-2025-2776 (CVSS score: 9.3) - An improper restriction of XML external entity (XXE) reference vulnerability in the Server URL processing functionality, allowing for administrator account takeover and file read primitives Both shortcomings were disclosed by watchTowr Labs researchers Sina Kheirkhah and Jake Knott back in May, alongside CVE-2025-2777 (CVSS score: 9.3), a pre-authenticated XXE within the /lshw endpoint. The three vulnerabilities were addressed by SysAid in the on-premise version 24.4.60 build 16 released in early March 2025. The cybersecurity firm noted that the vulnerabilities could allow attackers to inject unsafe XML entities into the web application, resulting in a Server-Side Request Forgery (SSRF) attack, and in some cases, remote code execution when chained with CVE-2024-36394, a command injection flaw revealed by CyberArk last June. It's currently not known how CVE-2025-2775 and CVE-2025-2776 are being exploited in real-world attacks. Nor is any information available regarding the identity of the threat actors, their end goals, or the scale of these efforts. To safeguard against the active threat, Federal Civilian Executive Branch (FCEB) agencies are required to apply the necessary fixes by August 12, 2025.