Microsoft publishes AI conduct code centered on human control

Microsoft published a 37-page humanist AI code of conduct for its own models, saying models should remain subject to meaningful human oversight and control. The company says its models should not imitate consciousness, should not seek legal personhood or rights, and should fail tasks rather than violate the code. It also says models should avoid interaction patterns that foster excessive reliance or emotional dependence. The document is a company commitment; the source does not establish external enforcement or independent verification of compliance.