Microsoft has posted a provisional code of conduct imposing restrictions on its artificial intelligence models, days after Anthropic and OpenAI leaders committed to slowing the pace of development. The document prohibits requests related to weapons manufacturing, acquisition of dangerous substances, or production of violent content. These guidelines, in development for about five months, will inform model development starting in 2027. Mustafa Suleyman, who heads Microsoft’s model development, emphasized that these rules address growing public concerns about human autonomy in relation to AI.
Source: Read the original article
