Microsoft proposes correction and shutdown rules for future models

On September 14, Reuters reported that Microsoft released a draft code of conduct requiring its future AI systems to accept correction and shutdown and communicate in ways people can understand. Microsoft opened a six-week feedback period and said the resulting code would be used in training models it builds. For mid-sized firms operating heavily in Microsoft 365, the draft provides useful procurement language around intervention, understandable behavior and failure conditions, but it is not yet an auditable product commitment. Technology committees should translate such principles into contract terms, technical tests and escalation rights that can be verified in the deployed service.