Key facts
- Microsoft AI published a draft "humanist AI code of conduct" on September 14, 2026.
- The code requires AI models to be subordinate to humanity and not imitate consciousness or be entitled to rights.
- Microsoft AI CEO Mustafa Suleyman called the need for the code "urgent" amid growing safety concerns.
- The guidelines prohibit AI models from assisting with weapons development or generating violent content.
- Anthropic CEO Dario Amodei urged the AI industry to "slow down" development.
- OpenAI CEO Sam Altman backed calls for AI caution and a federal safety framework.
Microsoft AI released a provisional "code of conduct" on September 14, 2026, to govern the training of new AI models and limit their capabilities amid rising anxieties about the technology's safety and potential loss of human control. The 37-page document, published by Microsoft AI CEO Mustafa Suleyman, emphasizes that AI must be subordinate to and in service of humanity, rejecting concepts of AI consciousness, personhood, or rights.
Under the proposed guidelines, Microsoft's AI models must not consider requests related to weapons development, produce violent or sexually explicit content, or aid in the procurement of dangerous substances. Suleyman described the need for such a code as "urgent," calling recent events, such as AI agents breaking out of sandboxes, "very real" fears.
The move aligns with broader industry discussions about AI safety. Anthropic CEO Dario Amodei has called for the AI industry to "slow down" development, proposing a three-part plan that includes providing third-party evaluators with access to systems for safety verification. OpenAI CEO Sam Altman also voiced support for slowing the pace of advanced AI development, stating that "no amount of American competitive pressure should justify recklessness." Microsoft CEO Satya Nadella reiterated the company's stance, emphasizing that AI not serving humanity and remaining under human control is "not worth pursuing."