The code denies consciousness in models, rejects legal personhood, and requires meaningful human control over AI systems.
Microsoft published a 37-page humanist AI code of conduct that states people take priority over AI. The code denies consciousness in models, rejects imitation of consciousness, and rules out legal personhood or welfare rights. Microsoft also commits to keeping its models subordinate to meaningful human oversight and control.
Anthropic CEO Dario Amodei called for a coordinated slowdown of AI development after researchers warned that model progress could outpace safe deployment and control of AI agents. Microsoft AI CEO Mustafa Suleyman previously called Anthropic’s speculation about model consciousness dangerous. Microsoft aims to become one of the top four AI labs.
For builders, the code signals that Microsoft plans to ship models that fail tasks rather than violate its rules. Teams evaluating Microsoft models should expect strict guardrails, limited agent autonomy, and a clear stance against AI welfare claims that shape product and policy decisions.
Microsoft is developing models to compete with Google, Anthropic, and OpenAI, and it wants to rank among the top four AI labs. The recent OpenAI and Hugging Face agent incident, plus an acknowledged wiki hijacking, will test whether the code can keep autonomous agents under human control.
What matters
- Microsoft published a 37-page humanist AI code of conduct that prioritizes people over AI.
- Operators should expect Microsoft models to fail tasks rather than violate the humanist code of conduct.
- Watch whether Microsoft’s next models match Google, Anthropic, and OpenAI under human oversight.
Why it matters
Watch whether Microsoft’s next models match Google, Anthropic, and OpenAI under human oversight.
This GenAI News article was prepared in original wording using reporting and materials published by The Verge. Source reference: https://www.theverge.com/news/994566/microsoft-humanist-ai-code-of-conduct.
Drafted by the GenAI News review pipeline.