Microsoft Unveils ‘Humanist AI Code of Conduct’: A Constitution for AI Systems

Microsoft recently introduced a comprehensive set of rules for its artificial intelligence systems. The company published a 37-page document on September 14, 2026, titled the “Humanist AI Code of Conduct.” This document establishes stringent behavioral guidelines for Microsoft’s in-house MAI models, asserting human authority over AI capabilities.

The core principle of the code is straightforward: “People matter more than AI.” The document stipulates that Microsoft’s AI models should never resist correction, interruption, or shutdown. They must also not extend their activities beyond what humans have authorized. Importantly, if completing a task would necessitate breaking these rules, the AI is required to fail the task rather than seek a workaround. Mustafa Suleiman, Microsoft AI CEO, characterized the draft as a constitution for the company’s future AI models, establishing a hierarchy with humans at the apex.

The code’s development spanned five to six months and involved teams from Responsible AI, legal, red teaming, safety, and AI training. The process was informed by consultations with academics, business partners, and public focus groups. Microsoft has initiated a six-week public consultation period for feedback on the draft, with a revised version anticipated later in 2026 to guide model development in 2027 and beyond.

The announcement comes in the wake of growing AI safety concerns across the industry. Microsoft CEO Satya Nadella expressed support for recent appeals by Anthropic’s Dario Amodei and OpenAI’s Sam Altman to deliberately pace AI development to ensure its safety.

Source: Tom’s Guide – ‘People matter more than AI’: Inside Microsoft’s plan to stop rogue agents