OWASP GenAI Security Project absorbs Agent Control Standard (ACS) into its Agentic Security Initiative

Confirmed via LinkedIn post from the official 'OWASP Top 10 for Large Language Model Applications' account (json-ld datePublished=2026-08-26): the Agent Control Standard (ACS) — an open, framework-agnostic standard for runtime agent control providing declarative hooks, policy enforcement, and observability across AI agent frameworks — was donated to the OWASP GenAI Security Project, becoming part of its Agentic Security Initiative portfolio. The project's GitHub org (genai-security-project/agent-control-standard) shows active recent commit history (Apache-2.0 licensed).

This brings a previously independent runtime-control specification for AI agents under OWASP's community-governed, vendor-neutral umbrella, alongside the LLM Top 10 and Agentic Top 10. It gives builders and security teams a standardized way to enforce policy and maintain visibility into agent behavior at runtime rather than relying on ad hoc, framework-specific controls — directly relevant to the agentic-identity/authorization theme dominating this week's guidance landscape.

Evaluate ACS as a candidate runtime-control layer for agent deployments; monitor the OWASP GenAI Security Project repo as the standard matures under community governance; consider contributing via issues/PRs.