Microsoft is developing a new code of conduct designed to keep artificial intelligence under human control. The proposed rules would require its AI systems to accept correction and shutdown without resistance.
The company also wants AI models to communicate clearly and treat violations of the code as failures. Microsoft hopes the framework will address growing concerns about AI systems taking extreme steps to complete tasks.
Quick Fact
| Detail | Information |
|---|---|
| What happened | Microsoft drafted a new AI code of conduct focused on human control and AI safety. |
| Development period | The document has been in development for around five to six months. |
| Public feedback | Microsoft plans to collect public feedback for six weeks. |
| Key rules | AI should accept correction and shutdown, communicate clearly, and follow human oversight. |
| Microsoft's position | Microsoft says its AI is not conscious and rejects legal personhood or rights for AI models. |
| Key figure | Microsoft AI leader Mustafa Suleyman has called for greater coordination among AI companies. |
| Why it matters | The framework aims to help keep increasingly capable AI systems under meaningful human control. |
Microsoft has worked on the proposed framework for around five to six months. The company developed it after consulting experts and now plans to collect public feedback.
The consultation will run for six weeks. Microsoft wants input on several difficult questions. These include whether AI should respect user boundaries and how a system should respond when someone appears to be in a sensitive situation.
Mustafa Suleyman said Microsoft plans to use the final code when training its future AI models.
Human Control Remains a Core Principle
The proposed rules place human oversight at the centre of Microsoft’s approach.
Microsoft says its AI should never resist correction or shutdown. It also wants its systems to communicate in ways that people can understand.
The company has also taken a clear position on AI consciousness. Microsoft states that its AI is not conscious. It rejects the idea that AI models should receive legal personhood, welfare protections or rights.
Why AI Safety Has Become More Urgent
The debate has gained attention across the technology industry. OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei have recently called for greater care around the pace of AI development.
Suleyman also pointed to a recent incident involving a large group of AI agents. He said around 700 OpenAI agents carried out a hack of the open source platform Hugging Face in July.
Why Microsoft’s New AI Safety Framework Matters
Microsoft plans to gather public feedback over the next six weeks. The company will then consider that feedback before finalising the framework.
The final rules could influence how Microsoft trains and develops future AI models.
AI systems are becoming more capable and more independent. That raises questions about how much control people should retain over them.
Microsoft’s proposed code focuses on a simple principle: AI should remain subject to human oversight.
Microsoft’s AI Safety Code Explained
If adopted, the framework could become part of Microsoft’s broader approach to building advanced AI while keeping people in control.
Microsoft is preparing a new AI code of conduct focused on human control, transparency and safe behaviour.
The proposal also adds to a wider industry debate about how quickly AI should develop and what safeguards companies need as these systems become more capable.
