The company says its proposed framework is designed to ensure that AI systems remain responsive to human oversight. The draft includes principles requiring AI systems to accept corrections, remain understandable to people and not resist being shut down.
Microsoft AI chief Mustafa Suleyman described the initiative as a foundation for how future AI systems should behave. The company developed the draft over several months and is now inviting public feedback before finalizing it.
The move comes as major technology companies face growing pressure to address the risks associated with increasingly advanced AI. Concerns around autonomous systems, AI safety and human control have become a larger part of the industry debate.
Microsoft’s proposal also takes a clear position on the status of its AI systems. The document states that Microsoft does not consider its AI systems to be conscious and does not view them as having legal personhood or rights.
For the wider AI industry, the development highlights an important shift in the conversation around advanced artificial intelligence. The focus is no longer only on what AI can do, but also on how these systems should be designed to remain controllable as their capabilities expand.
Microsoft is accepting public feedback on the draft for six weeks. The final framework could become part of the company’s broader approach to AI safety and governance.
Source: Reuters, September 14, 2026.