Microsoft sets provisional AI guidelines as industry leaders weigh safety slowdown

• Framework prioritizes maintaining human control over automated capabilities

US technology corporation Microsoft unveiled a draft governance framework on Monday to regulate its artificial intelligence systems, establishing safety parameters aimed at ensuring future models remain strictly under human supervision.

The draft rules, titled the Humanist AI Code of Conduct, establish core operating boundaries and values for future systems developed by the company.

The framework prioritizes maintaining human control over automated capabilities.

​​​​​​​Under the document, the software provider explicitly rejects the pursuit of all-purpose superintelligence that might evade oversight mechanisms, maintaining that systems must be designed to halt operations rather than breach conduct constraints.

It also explicitly stipulates that models will never resist human interruption, override, correction, or shutdown, and must always comply with a request to pause, redirect, cancel, or stop operations.

The publication opens a six-week public consultation window to gather external perspectives from ethicists, researchers, policymakers, and the public.

Following revisions toward the end of the year, Microsoft plans to implement the policy to guide model development in 2027 and beyond.

"We got feedback from people that they wanted to see even more explicit commitment to AI always working in service of people and not trying to replace them," Microsoft AI chief Mustafa Suleyman told television broadcaster CNBC.

Advanced artificial intelligence systems must remain firmly under human oversight while developers pursue deliberate pacing to ensure safe deployment, Microsoft Chairman and CEO Satya Nadella said Sunday.

The guidelines arrive days after executives from rival AI firms OpenAI and Anthropic agreed to slow the pace of development amid safety concerns.