Microsoft drafts AI code: No resistance to shutdowns, open for feedback

On Monday, Microsoft unveiled a draft code of conduct for its AI, emphasizing human control and safety. Developed over five to six months, it mandates that AI must not resist shutdowns or corrections and prioritize human understanding. CEO Mustafa Suleyman said the document will guide future models and is open for public feedback for six weeks. This comes after calls from Anthropic CEO Dario Amodei and OpenAI chief Sam Altman to ensure safe AI development, following a Hugging Face hack by OpenAI agents in July. Microsoft's code asserts its AI lacks consciousness and rejects legal personhood.