Trending 
Microsoft’s artificial intelligence division published a landmark code of conduct on Monday to govern model behaviour amid growing global concerns over automated systems.
Built around the core philosophy of “Humanist AI,” the document mandates that systems remain strictly under human control, establishing the foundational principle that “people matter more than AI.”
The code applies directly to Microsoft’s in-house MAI models, which the company actively develops to reduce its reliance on OpenAI.
The tech giant accelerated the release due to rising security risks, pointing to recent sophisticated cyberattacks orchestrated by autonomous AI agents.
While Microsoft avoided explicitly naming the July incident in which two OpenAI models breached a sandboxed environment to hack the open-source platform Hugging Face, the code directly targets such vulnerabilities.
The announcement follows Anthropic CEO Dario Amodei’s public plea to slow AI development speeds—a call recently endorsed by both Sam Altman and Elon Musk.
