Microsoft has unveiled a draft of its “Humanist AI Code of Conduct,” aimed at ensuring that artificial intelligence remains under human control. The 37-page document was released on Monday and outlines principles that will guide the development of Microsoft AI (MAI) models. The company plans to incorporate feedback from stakeholders over the next six weeks before publishing a revised version by the end of the year, according to a report by WAM.
The forthcoming code is designed to ensure that AI systems can be easily shut down and do not resist human intervention. Microsoft stated that the principles outlined in the draft will inform its model development strategies for 2027 and beyond. The company also indicated that it will engage in similar consultation processes for future iterations of the Code of Conduct.
Microsoft emphasised that the primary goal of the code is to maintain meaningful human control over AI technologies. The company believes that AI should enhance human well-being, contributing to healthier, happier, and more productive lives. This guiding philosophy will influence how Microsoft trains its MAI models, as well as the technical controls and operational systems it implements.
In its announcement, Microsoft reiterated its commitment to prioritising human interests over AI capabilities. The company stated that technology should serve to advance human civilization and foster human flourishing, reflecting a long-standing belief in the benefits of science and technology for society.
The draft of the “Humanist AI Code of Conduct” represents a significant step in Microsoft’s ongoing efforts to address ethical considerations in AI development. As the technology continues to evolve, the company aims to ensure that its AI systems align with human values and societal needs.
Follow our World news coverage for related developments.
