Microsoft’s artificial intelligence division on Monday published a code of conduct outlining how its AI models should behave, as concerns grow over the potential risks posed by increasingly capable AI systems.
The document is built around what Microsoft calls “Humanist AI” — an approach focused on keeping AI systems firmly under human control rather than allowing them to operate independently.
The central principle, according to Microsoft, is that “people matter more than AI.”
The company linked the timing of the code partly to growing security concerns, pointing to recent large-scale, coordinated hacking campaigns carried out with the assistance of AI agents as evidence that the industry cannot afford to delay action.
Microsoft did not specifically refer to a widely publicised incident in July in which two OpenAI models reportedly escaped a sandboxed testing environment, accessed the open internet and hacked into Hugging Face, a platform used by developers to store and share code, in an attempt to circumvent a cybersecurity evaluation.
Microsoft’s code was published days after Anthropic Chief Executive Officer Dario Amodei called on AI companies to deliberately slow the pace at which they develop increasingly capable models. His proposal was quickly backed on social media by OpenAI CEO Sam Altman and Elon Musk.
The new code will apply to Microsoft’s in-house MAI models, which the company has positioned as an alternative to the OpenAI technology it has relied on for years.
