Microsoft AI has published a draft of its Humanist AI Code of Conduct, an open-source technical manual that outlines operational constraints for model training and deployment. The document is the result of a six-week public consultation, during which Microsoft AI invited stakeholders to review and provide feedback on the draft code.
The proposed code defines system behavior, operational boundaries, and oversight protocols across various machine intelligence applications. It also builds upon Microsoft AI's humanist superintelligence framework, announced last November, which aims to develop artificial intelligence systems that are guided by a humanistic approach. This includes considerations for ethics, values, and human well-being.
The draft Humanist AI Code of Conduct is intended to provide a foundation for model development and deployment while minimizing the risk of unintended consequences. Microsoft AI plans to use the feedback from this public consultation to refine the code and make it more operationalizable in practice. The company has also stated that it will continue to develop its humanist superintelligence framework, with ongoing research and collaboration with stakeholders to ensure the ethical and beneficial development of AI systems.