Microsoft Unveils AI Code of Conduct Amid Safety Concerns

Microsoft has published a provisional code of conduct for its AI models to strengthen safeguards and address concerns over the risks posed by increasingly capable AI systems. The guidelines, developed over five months, require AI models to remain subordinate to human objectives and prohibit them from assisting with weapons development, generating violent or sexually explicit content, and imitating consciousness. Microsoft plans to gather public feedback before finalizing guidelines that will inform AI development starting in 2027.