Stocks
Microsoft's AI Code of Conduct Prioritizes Human Safety Over Model Autonomy
Microsoft has released a draft code of conduct for its homegrown AI models, emphasizing human control over model capability and autonomy.
The Humanist AI Code of Conduct prioritizes people's safety and well-being above task completion, even if it means a model failing to complete a request.
Microsoft's code draws on existing principles and frameworks, but goes further by addressing conflicts between user requests, operator policies, and Microsoft's own safety rules.
The document outlines strict limits on models' behavior, including prohibitions on using biological or chemical weapons, engaging in offensive cyberoperations, or participating in violent activity.