Microsoft AI released a first draft of its Humanist AI Code of Conduct for MAI models on September 14, 2026, and opened a six-week public consultation on how those systems should behave in development and deployment.
The company frames the document as a training and evaluation guide for Humanist AI: models that remain tools under human control, stay subordinate and contained, and never resist interruption, correction, or shutdown. Absolute constraints cover areas such as weapons of mass harm, child safety, and large-scale harmful manipulation, while enterprise partners retain room to configure defaults.
Microsoft says the draft grew out of internal Responsible AI, legal, red-teaming, and safety work plus consultations with academics, business partners, and community panels. Feedback is open for about six weeks; the drafting team says it will review comments, publish a summary of what changed, and release a revised version later in 2026 to guide MAI development from 2027 onward.
Independent coverage from The Verge and Reuters confirmed the launch timing and core tenets, including the premise that people matter more than AI and that models should not imitate consciousness or pursue legal personhood. The full draft and feedback channels are published on microsoft.ai.