Microsoft Bans AI Companions in Schools Under New Pact

Microsoft has agreed with the American Federation of Teachers (AFT) to adopt privacy and safety standards for its artificial-intelligence tools used in schools. The legally binding agreement requires third-party audits and will apply from 1 November to all schools with Microsoft contracts. Microsoft says it will not use student or educator data to train AI systems, except for a narrow student-protection purpose. Collected data cannot be sold, used for advertising or used for product development.

The rules also prohibit AI companions or features designed to encourage emotional attachment, dependency or engagement beyond the learning task, while requiring plain-language explanations for families about how the tools work.

The agreement follows growing concern over students using chatbots for schoolwork and personal or emotional support, and as New York City and Los Angeles introduced year-long pauses on student AI use while reviewing their approaches. The AFT described the deal as a potential industry standard, but online-safety group Fairplay warned that stronger privacy controls should not be taken as evidence that AI belongs in classrooms or is suitable at every age.

OpenAI and Anthropic said they were working with the AFT towards similar agreements. Google, whose Google Classroom is used by 150 million teachers and students worldwide, has not said whether it will sign. Experts said the framework could reduce the burden on individual schools to negotiate protections, but its broader impact depends on other major and smaller education-technology providers adopting comparable standards.