Autodesk taps Permiso Security to monitor AI agents across its cloud and workforce

What happened

Permiso Security Inc. launched new AI agent runtime security features to monitor AI-driven activity across both cloud and on-premises environments. Autodesk Inc. is the launch customer for this capability, using it to track AI agents operating within its cloud infrastructure and workforce. These features build on Permiso’s unified identity security platform, extending it to cover AI agents, which the company identifies as the fastest-growing identity type in enterprise environments.

Why it matters

The rapid adoption of AI-driven agents in businesses opens significant blind spots for security teams. Without continuous visibility into what these AI agents do, enterprises risk potential misuse, data leaks, or compromised workflows. Permiso’s extension of identity security to AI agents adds needed oversight that treats AI agents as distinct identities rather than just code or service accounts. This pressure on identity management practices forces security teams to rethink how agent behavior is audited and controlled in hybrid IT environments.

For Autodesk and similar organizations, the ability to monitor AI agents continuously allows them to detect anomalies or unauthorized AI activity sooner. It raises the bar on accountability and incident response, especially as AI adoption scales unpredictably. The move also highlights that AI agents are not just productivity tools but require security rigor commensurate with employee or service identities.

What to watch next

Watch for other enterprises adopting similar AI agent runtime monitoring to tighten security around AI automation. Vendors like Permiso may expand their identity frameworks, incorporating behavioral analytics and real-time policy enforcement specifically tailored for AI-driven processes. The evolution of AI agent security will affect compliance, operational security, and cloud governance approaches.

As AI agents proliferate across more enterprise functions, gaps in securing autonomous AI workflows will become riskier to ignore. Expect both new security tools and increased scrutiny from IT teams on how AI identities interact with sensitive data and systems.

AI Quick Briefs Editorial Desk