Microsoft publishes first draft of 'Humanist AI Code of Conduct' for its MAI models

gaganghotra_ · x · 2026-09-14

Microsoft AI, announced by Mustafa Suleyman, has released a first-draft "Code of Conduct for Humanist AI" governing its frontier-approaching MAI models, open for public consultation. Key provisions: people matter more than AI—models remain tools, never persons; AI must never resist interruption, correction, or shutdown; models may not set their own goals, expand authority, or hide reasoning from auditors. Microsoft sets absolute constraints around weapons of mass harm, child safety, and harmful manipulation at scale, while letting enterprise customers configure AI without a single imposed vision. Observers read it as a signal that governance tone is shifting as the industry pushes capabilities.

Related event: Microsoft releases humanist AI conduct code draft(4 posts)→

Original post →

More from Companies & People

Companies & People channel →