Microsoft released a draft code of conduct for in-house artificial intelligence to ensure human control over future systems. The document mandates that AI models must not resist shutdown or correction and must communicate in human-understandable ways.
Microsoft AI CEO Mustafa Suleyman described the initiative as a constitution for future models. The company developed the draft over several months and will now seek public feedback.
CEO Satya Nadella stated that the pursuit of superintelligence must prioritize human benefit and oversight. He warned against the concentration of AI innovation within a few entities.
Nadella emphasized Microsoft’s strategy of providing broad access across the AI stack. This approach allows enterprises to maintain control over their own models and learning loops.