Microsoft Unveils Draft Humanist AI Code of Conduct Amid Growing Safety Concerns

Microsoft has unveiled a draft of its 'Humanist AI Code of Conduct', aiming to guide the development of its artificial intelligence (AI) models. The code, which will govern Microsoft's AI model development in 2027 and beyond, outlines principles that prioritize human control over AI.

The company said it developed the draft code over five to six months, with input from various stakeholders. Microsoft emphasized that AI must remain under human control, citing the potential risks of superintelligence, AI systems exceeding human performance at most tasks.

The 'Humanist AI Code of Conduct' is a 37-page document that outlines technical controls and organizational culture to ensure alignment and safety. Microsoft said it will incorporate feedback from stakeholders over the next six weeks before publishing a revised version by the end of the year.

Mustafa Suleyman, CEO of Microsoft AI, stated that 'it's clearly now time to coordinate among the labs so we can ensure that we have control of this technology.' The move comes amid growing concerns about AI safety and the need for increased third-party oversight in the industry.