jetstream

JetStream Security enables on-demand shutdown of compromised AI agents JetStream Security has announced the release of an AI Kill Switch that allows organizations to shut down compromised AI agents on-demand without impacting other AI operations. This new control plane for AI agents solves the inability to stop a single agent that falters, begins overspending, or needs to be taken offline for compliance reasons without affecting other agents. JetStream’s AI Kill Switch solves a problem many businesses have not faced up to this point: AI systems are becoming increasingly capable and as evidenced by the recent OpenAI/Hugging Face incident, can cause significant collateral damage as they seek to attain their goals in ways that we cannot predict. This new reality means that every organization needs to have the ability to continuously monitor its AI systems and have the ability to precisely halt malicious actions using a kill switch. “Not being able to stop a runaway AI system without impacting other production systems is one of the chief reasons companies hold back on AI,” said JetStream Security CEO, Raj Rajamani. “Only 17% of organizations are running AI at production scale, held back by a lack of trust and control rather than by the models. The AI Kill Switch™ failsafe frankly reduces our customers’ anxiety. If one system goes wrong, they stop that one and everything else keeps running.” To provide this level of precision, JetStream factors together the individual who invoked the action, the agentic identity that carries it out, the intent of the system as inferred from its AI Blueprint and historic actions, and the individual product owner accountable for that agent. Access entitlements are mapped out in JetStream AI Blueprints so that every agentic action is traceable back to a responsible owner, enabling the Kill Switch to revoke exactly the right access and shut off the correct components in one pinpointed action. Because each agent’s authority is carried by its own entitlements, the AI Kill Switch affects the single agent alone, and these actions are reversible. JetStream furthers the building of trust by recording every action made by every actor with a corresponding timestamp that gives organizations a defensible account for post-incident review. That record matters beyond the incident itself. It gives companies something boards, auditors, and a growing body of AI regulation increasingly ask for: proof that a person can halt an AI system on demand, and clear evidence of who stopped it and why. “Regulators and standards bodies have quietly converged on something most enterprises haven’t caught up to which is if you can’t stop an AI agent, you don’t control it. The EU AI Act now requires a way to bring a high-risk system to a safe halt, and both the NIST AI Risk Management Framework and ISO/IEC 42001 expect organizations to be able to disengage or shut down AI that operates outside its approved boundaries. A stop button is no longer a ‘nice-to-have’, it’s the line between AI you govern and AI that’s governing itself.”

Kevin Mandia, who previously founded Mandiant and sold the incident response and threat intelligence firm to FireEye in a $1 billion deal in 2014 before its later $5.4 billion acquisition by Google, has launched a new cybersecurity startup. San Francisco based Armadin announced a massive Seed and Series A Funding round of $189.9M in what is effectively its public launch (the firm has simultaneously published a blog by Mandia titled Introducing Armadin). Armadin uses AI-powered red teaming to find and exploit weaknesses in the same way that attackers attack them. The funding was led by Accel, with participation from Google Ventures, Kleiner Perkins, Menlo Ventures, In-Q-Tel, and follow-on investment from 8VC and Ballistic Ventures — and is claimed to be the largest seed and series A funding in cybersecurity history. The firm was first announced quietly in late 2025 with an initial seed of $24 million. Now, with the formal launch, Mandia’s co-founders have been named as Travis Lanham (CTO), Evan Peña (chief offensive security officer), and David Slater (chief architect). Mandia is CEO. Armadin can be described as a red team on steroids. The steroids are AI. “I believe within the next few years virtually all cyberattacks will be AI-based – swarming, tailored, and relentless,” writes Mandia in his blog. “They will be untethered to human limitations and capable to execute on a scale we have never witnessed before.” Armadin intends to fight fire with fire. “In a world of machine-speed attacks, defense must become autonomous. You cannot have a human in the loop for every defense decision and expect to win,” he adds in the ‘launch’ document. We are building the most formidable offense to give organizations the greatest defense. It’s important to national security.” The ‘formidable offense’ comes from a team of specialist red teamers and AI researchers and engineers. The intent is to provide a platform that can perform AI-directed offensive security as fast as attackers can attack – to close or at least reduce the traditional agility gap – with its own autonomous, agentic attacker swarm. “Before Armadin, you could not put a nation state level adversary inside every network 24/7,” said Lanham. “We’ve built the ultimate attacker – it doesn’t just follow a script, it reasons and learns as it swarms your defenses. We train our models and build agents to the standards of a world-class red team with safety at the foundation and unleash them to identify exploitable risk at machine speed.” Traditional red teaming cannot cope with the dawning age of AI-driven attacks. Armadin intends to solve this with AI red teaming. Related: Palo Alto Networks Founder Nir Zuk Unveils New Startup Cylake Related: AI Security Firm JetStream Launches With $34 Million in Seed Funding Related: Fig Security Launches With $38 Million to Bolster SecOps Resilience Related: Aisy Launches Out of Stealth to Transform Vulnerability Management

Developer-focused data security company Evervault on Thursday announced raising $25 million in Series B financing to expand its infrastructure and team. The round was led by Ribbit Capital with participation from Index Ventures, Sequoia Capital, Kleiner Perkins, and Operator Partners. The company has raised $46 million in total funding, with the latest capital set to support product development, engineering and product team growth, and expansion of its encryption system. Founded in 2019, Evervault provides a platform that allows organizations to encrypt and orchestrate sensitive data without accessing it in plaintext. The platform integrates encryption with features like authentication, token management, and data enrichment, particularly for handling card payments. It includes tools such as a relay proxy for securing network requests, secure environments for running code on encrypted information, user interface elements for collecting payment details, and isolated containers for processing sensitive workloads. “Most compliance frameworks assume sensitive data will exist in plaintext somewhere, but with automated, high-velocity data exchange, that’s a liability,” said Shane Curran, founder and CEO of Evervault. “At Evervault, we believe sensitive data should be treated like hazardous material. Systems must be designed so it isn’t touched in the first place. We’re building the Internet’s trust layer, embedding encryption directly into the application architecture so it stays encrypted by default, not by policy alone,” Curran added. Related: Reclaim Security Raises $20 Million to Accelerate Remediation Related: AI Security Firm JetStream Launches With $34 Million in Seed Funding Related: Fig Security Launches With $38 Million to Bolster SecOps Resilience

Reclaim Security on Wednesday announced raising $20 million in a Series A funding round that brings the total raised by the company to $26 million. The startup’s most recent investment round was led by Acrew Capital, with additional support from QP Ventures and Ibex Investors. Founded in 2024, US-based Reclaim has built an autonomous platform that helps organizations identify security risks and address them fast, to cut remediation gaps. The solution relies on an AI Security Engineer designed to identify exposures and resolve them at scale. It relies on a simulation engine named PIPE (Productivity Impact Prediction Engine) that predicts the impact of any change on applications, workloads, productivity, and business processes before deployment, to eliminate downtime and operational disruptions. The approach, Reclaim says, allows organizations to prioritize exposures most likely to be exploited, deploy remediations safely, reduce remediation times to minutes, and eliminate manual work. Through the analysis of attack techniques’ impact on environments, security tools’ response, and operational impact, exploitable pathways are eliminated safely at scale, Reclaim says. The startup will use the new investment to expand its engineering team, deepen integrations, and accelerate go-to-market efforts in North America and Europe. “Security tools are excellent at explaining why something is risky. What they don’t do is make remediation safe and practical. The real breakthrough isn’t more prioritization, it’s removing risk without breaking the business. Reclaim does exactly that, and that’s why it matters,” said Acrew Capital founding partner Mark Kraynak. Related: AI Security Firm JetStream Launches With $34 Million in Seed Funding Related: Fig Security Launches With $38 Million to Bolster SecOps Resilience

AI governance startup JetStream Security has launched with $34 million raised in an oversubscribed seed funding round led by Redpoint Ventures, with additional support from CrowdStrike Falcon Fund and angel investors. Founded by security veterans from Attivo Networks, CrowdStrike, Cohesity, Dazz, McAfee, and SentinelOne, US-based JetStream aims to help enterprises securely deploy generative and agentic AI tools in production. The startup has built an AI governance platform that leverages AI Blueprints, described as dynamic, system-generated graphs providing real-time visibility into how AI operates across enterprise environments. The blueprints were designed to map the relationships between AI agents and their models, the accessed data, the tools they use, and the identities behind them. They leverage real runtime behavior tracking, flag deviations, track workflow costs, and can be updated through authorization workflows. They make behavior and costs visible, governable, and attributable. JetStream aims to accelerate the adoption of AI within enterprise environments by helping businesses understand and control AI agents’ operations across their entire stack. The startup plans to expand its engineering, product, and go-to-market teams. “AI is moving faster than most organizations can manage. With AI Blueprints, we give teams a clear, practical way to understand what their AI is doing, manage risk in real time, and move from experimentation to production with confidence. Our goal is simple: help companies scale AI responsibly, without slowing innovation,” said JetStream CEO and co-founder Raj Rajamani. Related: Fig Security Launches With $38 Million to Bolster SecOps Resilience Related: Gambit Security Emerges From Stealth With $61 Million in Funding