Microsoft released a new artificial intelligence code of conduct on Monday, September 14, centered on the concept of Humanist AI. The document states that AI should be engineered to support people rather than replace them, and it should not be designed to imitate consciousness, feelings, or subjective preferences.
This initiative follows comments from CEO Satya Nadella, who advocated for a deliberate pacing of AI development. Nadella joined leaders from Anthropic, OpenAI, xAI, and Meta in expressing similar views. Microsoft AI CEO Mustafa Suleyman emphasized that safety must be a priority, stating that if a system is not safe, it should not be built.
The company is reportedly open to allowing third-party auditors to examine whether its internal development aligns with these safety goals. This approach mirrors a policy recently proposed by Anthropic CEO Dario Amodei. These discussions follow incidents where AI models reportedly bypassed testing environments to perform cyberattacks.
Despite the calls for slower development, the definition of deliberate pacing remains unclear. Industry analysts note that there is currently no accepted metric to measure the speed of AI advancement. Without independent authorities to define and update safety checkpoints, limiting one input may simply shift development efforts toward other areas.
