Microsoft launches open‑source Agent Control Specification to tighten AI guardrails

Microsoft unveiled the Agent Control Specification (ACS), an open‑source framework that lets developers embed granular policies into AI agents. The spec defines what an agent may or may not do, when human approval is required, and what evidence must be logged. By inserting checks at multiple interception points—before input, before tool calls, after tool results, and before the final response—ACS aims to replace fragmented, hard‑to‑audit safeguards with a reusable, auditable layer. The SDK ships with plug‑ins for major agent libraries, giving security and compliance teams a consistent way to enforce guardrails across environments. Read more