Microsoft Proposes Code of Conduct for AI Development Amid Industry-Wide Safety Concerns
Microsoft has proposed limits on its AI development with a provisional 'code of conduct' aimed at preventing AI from getting out of control. The code, published by Microsoft's CEO of AI Mustafa Suleyman, outlines guidelines for training new AI models, including not considering requests related to weapons development or producing violent or sexually explicit content.
The move comes as the industry grapples with concerns over AI security and safety. OpenAI's CEO Sam Altman has warned that the rapid pace of progress could go 'very badly' if not managed carefully. Meanwhile, Microsoft's Satya Nadella emphasizes that AI should be built to serve humanity and under human control.
The code also includes provisions for third-party evaluation and monitoring of AI systems, but some experts have expressed skepticism about its effectiveness without global cooperation. Oliver Yonchev, co-founder and COO of Potentially AI, warns that the biggest labs could write rules that protect their own position if not implemented with international agreement.