Microsoft drafts feel-good AI model guidelines and wants your input

Microsoft's AI guidelines include aspirational goals but lack stated consequences for non-compliance.

Microsoft has released draft guidelines for its AI model development, called the Humanist AI Code of Conduct, and is seeking public feedback. The company aims for its AI models, referred to as MAI models, to help people live healthier, happier, and more productive lives, while ensuring humans retain meaningful control. These principles are intended to govern the training, technical controls, and monitoring of future MAI models, even those approaching 'superintelligence.' The document acknowledges a gap between current model behavior and these aspirational goals, stating it is a 'north star' and not a guarantee of present performance. Microsoft is accepting public input on these guidelines until October 25, 2026. The company also recently announced layoffs affecting around 4,800 employees, with assurances that these roles are not being replaced by AI.

Microsoft's Humanist AI Code of Conduct, while aspirational, appears to be a public relations effort rather than a binding commitment. The document explicitly states it is not a guarantee of present-day performance, leaving a significant gap between its lofty goals and current realities. This approach, devoid of specified repercussions for internal failure to adhere to its principles, echoes historical instances of corporate self-regulation falling short.

The company's timing for soliciting feedback on AI governance is notable, especially as competitors like Anthropic and OpenAI have called for regulatory intervention. By presenting its 'aspirational' guidelines and seeking public input, Microsoft positions itself as a proactive, responsible player in the AI space, potentially influencing future regulatory frameworks without facing immediate constraints.