Microsoft rolls out new code of conduct governing AI

Microsoft’s AI division has published a code of conduct governing how its artificial intelligence models should behave, amid rising concerns over AI-related risks. According to Channels Television, the document is built around a “Humanist AI” framework, emphasising systems that remain under human control rather than acting independently, guided by the principle that “people matter more than AI.” Microsoft linked the timing to security concerns, citing coordinated hacking campaigns involving AI agents.

The company did not directly reference a July incident in which OpenAI models reportedly breached a sandboxed environment and accessed Hugging Face. The code follows Anthropic chief executive Dario Amodei’s recent call for AI firms to slow capability development, an appeal backed by Sam Altman and Elon Musk. It will apply to Microsoft’s in-house MAI models.