AI Runtime Security & Agentic Zero Trust
Key Questions
What are the main challenges with AI agents in zero trust environments?
Federal systems face significant hurdles managing 7.4M authentication events daily from AI agents, alongside endpoint coverage gaps of 12.7% reported by Axonius and Ponemon. Key themes include the confused deputy problem and the need for identity analytics in non-human identity governance.
How does ISO 42001 support AI accountability and governance?
ISO 42001 provides a framework for implementing structured AI governance to ensure accountability, such as explaining decisions made by AI systems like loan denials. It helps organizations address risks in AI runtime security and agentic environments.
What new tools and regulations are emerging for AI runtime security?
Recent developments include Permiso Risk Score Engine, BeyondTrust NHI Governance module, Vectogate, Keeper Security reaching $225M ARR, and California SB 53 AI regulation. Acquisitions like Barracuda acquiring Evo Security and Cisco's NHI moves, along with the CoSAI framework, highlight the rapid evolution in this domain.
Dominant domain with rapid developments: Google Agent Identity (SPIFFE), Axonius/Ponemon endpoint coverage gap (12.7%), Cisco NHI acquisitions, federal zero trust challenges with AI agents (7.4M auth events/day). New: Permiso Risk Score Engine, BeyondTrust NHI Governance module, ModelCop, Vectogate, Keeper Security $225M ARR, Barracuda acquires Evo Security, NetApp CPO on data-layer zero trust for agentic AI, practical guides on implementing AI securely. ClearML's six-layer security stack and ISO 42001 as a certifiable AI management system add structured frameworks. CoSAI framework, confused deputy problem, and identity analytics for NHI governance remain key themes.