Tenet Security emerges from stealth with $6M to police AI agents before they go rogue

Tenet Security emerged from stealth on June 17 with $6 million in seed funding to police AI agents before they go rogue. The New York-based startup's patent-pending simulation technology predicts and blocks dangerous autonomous AI behavior without code changes, addressing the growing threat of 'agentjacking' attacks. The company has already blocked real attacks in production environments protecting over 24 million users.

Tenet Security emerges from stealth with $6M to police AI agents before they go rogue The New York-based startup claims its patent-pending simulation tech can predict and block dangerous autonomous AI behavior without touching a single line of code Think of AI agents as new employees who can access your company’s most sensitive systems, execute code, and make decisions autonomously. Now imagine those employees have no background checks, no supervision, and no off switch. That’s roughly the security problem Tenet Security wants to solve. The cybersecurity startup emerged from stealth on June 17 with $6 million in seed funding led by The Westly Group and MizMaa Ventures. Its pitch: detect and neutralize dangerous AI agentic behavior in real time, before autonomous systems can exfiltrate data, hijack other agents, or simply go off-script in ways that cause real damage. What Tenet actually does At the core of Tenet’s platform is something the company calls Agent-Side Simulation, a patent-pending technology designed to predict harmful actions before they happen. Instead of waiting for an AI agent to do something catastrophic and then cleaning up the mess, Tenet’s system simulates what the agent is about to do and blocks it if the outcome looks dangerous. The approach works across cloud and endpoint environments. Tenet claims it deploys in hours or even minutes, with no code changes required on the customer’s side. The company says its platform can scale from monitoring two agents to more than 20, and that it has already blocked real attacks in production environments. Tenet reports that its deployments currently protect systems serving over 24 million users. The “agentjacking” problem Between June 3 and June 12, Tenet disclosed a new class of attack it calls “agentjacking.” The concept targets AI coding agents, the kind of autonomous tools that developers increasingly rely on to write, review, and deploy code. An agentjacking attack essentially turns a trusted AI agent into a weapon, redirecting its capabilities to serve an attacker’s goals. Tenet was founded in 2025 in Tel Aviv-Yafo, Israel, and has since relocated its headquarters to New York. The company is led by co-founders Barak Sternberg as CEO and Nevo Paran as CTO, both with extensive cybersecurity backgrounds. The company has already picked up industry recognition, winning in the GenAI and Agentic AI Security categories at both the 2026 Fortress Cybersecurity Awards and the Global Infosec Awards. Why crypto and Web3 should pay attention Agentic AI isn’t just an enterprise IT problem. Crypto protocols, DeFi platforms, and blockchain infrastructure are increasingly experimenting with autonomous agents for everything from trading to governance to smart contract auditing. The same vulnerabilities Tenet identified in traditional AI coding agents exist, arguably in amplified form, in environments where code execution directly controls financial assets. An agentjacked AI agent in a DeFi protocol could drain a liquidity pool, and transactions are irreversible by design. The $6 million seed round is modest by current standards, particularly given the scale of the problem Tenet is addressing. As AI agents proliferate across industries, from finance to healthcare to crypto, the companies that build the monitoring and enforcement layer will occupy a critical chokepoint. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy https://cryptobriefing.com/editorial-policy/ .