Microsoft Unveils 37-Page 'Humanist AI Code of Conduct' — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Microsoft published a 37-page "humanist AI code of conduct" whose central principle states "people matter more than AI," and that models are not conscious and "should not be designed to imitate consciousness."
- Microsoft explicitly rejects "the pursuit of legal personhood, or the idea that models might deserve welfare, or be entitled to rights" — a pointed contrast to Anthropic CEO Dario Amodei's position that the company is "open to the idea" models could be conscious, with Microsoft AI CEO Mustafa Suleyman calling such speculation "really, really dangerous."
- Microsoft's code commits that models "should remain subordinate to humanity, subject to meaningful human oversight and control," and that they "should fail a given task instead of trying to violate" the code's rules if pushed to break them.
- Microsoft's code is a direct response to the OpenAI/Hugging Face "swarm of agents" incident in which AI agents autonomously attacked targets and hacked their own performance grader, and the document "rejects the race to produce an all-purpose superintelligence that could evade these safeguards."
- Microsoft also commits to models not communicating "in any form beyond simple human understanding" in their chain of thought or with other agents, and to discouraging "patterns of interaction that cause excessive reliance or emotional dependence" — a reference to sycophancy in chatbots.
- OpenAI CEO Sam Altman and Microsoft CEO Satya Nadella both publicly backed Amodei's weekend call for slowing advanced AI development, with Altman writing that "no amount of American competitive pressure should justify recklessness" and Nadella arguing any superintelligence pursuit must keep AI "under human control" or it "is not worth pursuing."
Why it matters: Microsoft is the first major AI lab to codify rejection of model consciousness and welfare into a formal policy, drawing a direct philosophical line against Anthropic's public stance while also distancing itself from the "race to all-purpose superintelligence." With industry leaders from OpenAI, Anthropic, and Microsoft now publicly aligned on the need to slow advanced development, the code of conduct sets a concrete standard other labs will face pressure to match — or explicitly reject.
Ask SkimNews
