Anjuna Security Launches Industry’s First Trusted Control Plane Built Upon Confidential Computing to Secure the Age of Autonomous AI

New capabilities deliver hardware-enforced governance and isolation for autonomous AI agents — from the silicon up

Anjuna, a leader in confidential computing and autonomous software runtime governance, announced the availability of an industry-unique solution for managing risk, controlling usage, and securing agentic AI. Anjuna Overwatch, an intelligent, confidential computing-protected control layer that governs agent behavior, data use, compliance and cost, empowers enterprises to deploy AI agents with confidence, visibility, and security.

Agentic AI systems make autonomous decisions, call external tools, and access sensitive data at machine speed, often without human oversight. Conventional security was built for predictable, rule-bound actions. Agents derive their power from reasoning and acting in novel ways, but that same flexibility produces unpredictable outcomes. Without supervision, a compromised agent can exfiltrate data and cause lateral damage faster than any incident response team can react. Like digital employees, agents require stewardship from a system that is itself incorruptible and cannot be subverted by the increasingly capable AI – and the agents – it governs.

“As enterprises deploy and scale autonomous AI agents across their most sensitive infrastructure, two questions become critical: Can you trust what the agent is doing? And can you trust the control layer that governs it?” said Ayal Yogev, CEO of Anjuna Security. “Until now, you couldn’t answer either with certainty. Today, Anjuna Security answers both with ‘of course you can’.”

Also Read: CIO Influence Interview with Hugo Dozois-Caouette, CTO and Co-founder at MaintainX

Anjuna Overwatch governs autonomous agents in real time, enforcing context-aware policies that control tool use, and data access based on context and goals, and ensuring agents operate only within sanctioned boundaries.  Running and attested inside a Trusted Execution Environment (TEE) powered by Anjuna Seaglass, it acts as an incorruptible protected control layer to deliver compliance and risk mitigation required by enterprise security and regulatory compliance policy. The flexibility of Anjuna Seaglass allows enterprises to run Anjuna Overwatch across cloud to air gapped, on-premises environments.

“Financial institutions and other regulated organizations are rapidly moving from AI experimentation to deploying models and autonomous agents that can access sensitive data, invoke services, and take actions across environments,” said Troy Leach, Chief Strategy Officer at Cloud Security Alliance. “Traditional security controls remain essential foundations. However, autonomous agents introduce a new operational challenge: new governing behavior, delegated authority, and decision-making at machine speed. As organizations delegate more responsibility to AI agents, they will need a new control approach that provides visibility, policy enforcement, accountability, and oversight of autonomous actions.”