Microsoft Agent Control Specification Brings Deterministic Runtime Enforcement, Leaving the Authorization Decision Upstream, June 2026

GOVERNANCE IMPLICATION

ACS is a genuine advance in agent governance and worth acknowledging plainly: it moves control from probabilistic instruction, telling a model what not to do, to deterministic enforcement at defined checkpoints, expressed as portable, versionable, auditable policy. That is real progress on the enforcement layer. It also makes the layer above it sharper rather than smaller. Sarah Bird's framing in the announcement is that written policies do not translate into working runtime controls, and ACS is the translation. What ACS assumes as input is the policy itself, the decision about what a given agent is permitted to do and who is accountable when that permission proves wrong. That authorization decision sits upstream of all five checkpoints. A better enforcement mechanism does not answer who authored the rule or who owns its outcome. When the lock on the door gets stronger, the question of who holds the key, and who decided the door should exist, gets more important, not less. ACS gives the Authorization Coverage Lifecycle a concrete enforcement target, and it leaves the authorization decision exactly where it was.

THE GOVERNANCE QUESTION

ACS enforces your agent control policy deterministically at five checkpoints. Who in your organization authored that policy, who is accountable for the authorization decision it encodes, and where is that decision recorded before it becomes a YAML manifest?

REGULATORY RELEVANCE