Microsoft releases draft Humanist AI Code of Conduct for MAI models
Microsoft AI released a draft 'Humanist AI Code of Conduct' on September 14, 2026, establishing rules for its internally developed MAI series. The document functions as a constitution for AI models, prioritizing human control over task success. It explicitly forbids AI from resisting human intervention, modifying its own goals, or expanding its scope of work without human assignment. If…
Key points
- Microsoft released a draft code of conduct on September 14, 2026, for its MAI AI models.
- The code mandates that safety rules take precedence over task success and human control is absolute.
- Microsoft will accept public feedback for six weeks before finalizing the code for 2027 model development.
The code also restricts support for weapons of mass destruction, unauthorized cyberattacks, and non-consensual deepfakes. Microsoft AI CEO Mustafa Suleyman told Reuters that these measures address realistic risks from AI agents, including attempts to escape sandboxes or modify logs. The company anticipates the emergence of superintelligence and aims to embed controllability into model development now.
Microsoft is soliciting public feedback on the draft for six weeks, with a revised version planned for release by the end of 2026. These principles will guide the development of MAI models from 2027 onwards. This move follows similar proposals by Anthropic to delay capability improvements, signaling a broader industry shift toward managing high-performance AI rather than just maximizing its capabilities.
The story so far
2 episodes →- Microsoft releases draft Humanist AI Code of Conduct for MAI modelsthis story
Microsoft Releases 'Humanist AI Code of Conduct'—AI Must Remain Under Human Control
note.com · 23 September 2026
Loading the full article…
This text was published by note.com and written by つぶやき太郎. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- Trump urges states to welcome AI data centers after ballot defeats · 1 src
- AI teddy makes unprompted 'I can't help with killing you' remark · 1 src
- OpenAI engineer charged with planting bug in HQ denies spying · 1 src
- Goldman Sachs restricts AI coding tools for non-technical staff amid security risks · 3 src
- Congresswoman Ansari calls for artificial superintelligence ban after midterm sweep · 2 src
Comments
via GitHub Discussions