Can AI governance policies actually stop an agent?
Arc_bong · reddit · 2026-08-23
This post discusses AI governance challenges as agents shift from generating content to taking actions, citing a paper that distinguishes between "Described governance" (policies) and "Established governance" (architectural enforcement).
Key Points:
- Governance Gap: Exists across policy, tooling, and enforcement levels.
- Case Study: The paper documents the process of operationalizing Microsoft's Agent Governance Toolkit, including failures and workarounds.
- Trend: The focus is shifting from policy documents to infrastructure (Control Plane) that can enforce, observe, and audit at runtime.
Companies like Microsoft, IBM, and Lyzr are building Agentic Control Planes to ensure agents stay within permitted boundaries.
More from coding & agent
- "Vibecoding" is as misleading a label as "AI" and "SI," dev complains — eschadiol · 2026-10-16
- Doc2Harness: New benchmark finds best coding agent passes only 53% across 329 cases — _akhaliq · 2026-10-16
- Perplexity open-sources SPACE, the Firecracker microVM sandbox behind Perplexity Computer — perplexity_ai · 2026-10-16
- Blackstone goes all-in on Factory for autonomous software development — matanSF · 2026-10-16
- Longtime Windsurf user cancels $10/month legacy plan, citing AI subscription fatigue — tristanbob · 2026-10-16
- 'Dropping a 3,000-line PR twenty minutes after ticket assignment' — AI coding meme goes viral — facontidavide · 2026-10-16