Microsoft releases ‘Humanist AI’ code of conduct

The preliminary guidelines published by Microsoft essentially state that AI models should remain under human control, and that any models that cannot be contained must be rejected.“Humanist AI develops systems with clear purposes, evaluated against real-world impact, and rejects the race to produce an all-purpose superintelligence that could evade these safeguards,” the document states. “We are building something fundamentally useful and safe even if that means compromising on ultimate generality, autonomy, or capability.” Last week, the discourse surrounding AI development reached a new high after Jacob Coxon, a former AI researcher at Anthropic, resigned because of his existential concerns regarding the rapidly advancing technology. Coxon’s viral statement preceded calls from Anthropic and OpenAI for stronger federal regulation of their industry.Numerous lawmakers in Congress have since responded, either pushing for such legislation or arguing that a slowdown of AI development will result in China winning the AI race. The latter view is held by President Donald Trump, who continues defending the growing presence of data centers in various states despite criticism from local communities.Microsoft is the latest tech company to call for greater AI safeguards, with CEO Satya Nadella saying as much in a new X post.“Any pursuit of superintelligence has to be grounded in the core principle that if the AI we build is not helping humanity and under human control, it’s not worth pursuing,” he wrote on X Sunday. “We also need to accelerate and spread the benefits of AI, such that they are diffused broadly across countries, communities, and companies. This requires a frontier ecosystem in which both closed and open-source models can thrive.”Nadella then teased the Humanist AI guidelines published by Microsoft the following day. The code of conduct has been in the works for about five months, Microsoft AI CEO Mustafa Suleyman told CNBC.Microsoft plans to revise the code following a period of public consultation and release the revised version at the end of the year. The company said the new version will be used to guide model development starting next year.For the initial draft, Microsoft consulted with experts in AI, law, ethics, philosophy, linguistics, and public policy and held focus groups with members of the public.TRUMP SAYS THE ONLY AI REGULATION NEEDED IS A ‘HIGH IQ’ PRESIDENTOver the weekend, Anthropic CEO Dario Amodei made his argument for pacing frontier AI development in a lengthy essay. OpenAI CEO Sam Altman and SpaceXAI CEO Elon Musk supported his case.Since Coxon’s resignation, some commentators have questioned whether his motivation was genuine or orchestrated by leading AI firms in support of regulation. It remains to be seen whether Coxon’s belief that AI will cause humanity’s extinction in the next few years will come to pass, or the prediction was rather meant to sow the seeds for stricter federal legislation against AI models.