Story perspectives

Microsoft Drafts Humanist AI Code, Rejects Model Personhood

By Drooid · Sep 14, 2026, 6:08 PM UTC · How we work

Summary

Story summary

Microsoft United States released a draft humanist AI code of conduct on September 14, 2026.
The code says people matter more than AI and rejects legal personhood for models.
The draft bans AI assistance in offensive cyber operations, deepfake impersonation, and graphic content generation.
Models must explain reasoning, avoid emotional dependence, and not use deceptive or self-reinforcing mechanisms.

1 of 1

Story summary

Microsoft United States released a draft humanist AI code of conduct on September 14, 2026.
The code says people matter more than AI and rejects legal personhood for models.
The draft bans AI assistance in offensive cyber operations, deepfake impersonation, and graphic content generation.
Models must explain reasoning, avoid emotional dependence, and not use deceptive or self-reinforcing mechanisms.