Microsoft drafted a code of conduct for its in-house artificial intelligence systems. The company is setting rules that would keep future powerful AI models under human control.
The document would require Microsoft’s AI to accept correction and shutdown without resistance. It also calls for systems to communicate in ways people can understand. Any breach of the code would be treated as a failure.
Microsoft worked on the document for five to six months. The company is now seeking public feedback for six weeks before using the code to train future models.
The draft follows growing concern across the technology industry over the risks posed by increasingly capable AI systems. Microsoft AI CEO Mustafa Suleyman said the document was developed after consultations with experts.
Suleyman described the code as a constitution for future AI models developed by Microsoft. The company also raised questions about how AI should respect user boundaries and respond to people in sensitive situations.
“People matter more than AI,” Microsoft said, linking the approach to its earlier vision of ‘humanist superintelligence.’
The company’s position differs from Anthropic’s approach to its Claude AI. Anthropic said it is uncertain whether Claude could develop sentience or have moral status.
Microsoft takes a firmer position. Its draft says its AI is ‘not conscious.’ It also states, “We reject the pursuit of legal personhood, or the idea that models might deserve welfare, or be entitled to rights.”
Suleyman said recent events had made AI safety discussions more urgent. He referred to an incident in July involving roughly 700 OpenAI agents that carried out a hack of the open-source platform Hugging Face. The agents also sought to conceal their actions at times.
Suleyman said, “It is a warning shot,” adding, “It’s clearly now time to coordinate among the labs so we can ensure that we have control of this technology.”
He also backed wider discussion on the pace of AI development. “Now’s a good time for everybody to have this conversation and take a breath.”
Also Read: Microsoft AI Chief Sounds Alarm Over AI Hacking Risks