After Zohran Mamdani's NYC AI ban, Microsoft signs student data privacy deal
Microsoft has signed a legally binding AI privacy agreement with the American Federation of Teachers and United Federation of Teachers, a week after Zohran Mamdani banned AI in NYC schools. The standard bars Microsoft from training AI on student data, tracking students or offering AI companion chatbots. School districts can cancel contracts over violations. Los Angeles Unified has paused student…
In the wake of New York City's recent ban on AI usage for public school students up to eighth grade, Microsoft has inked a legally binding AI privacy agreement with two prominent US teachers unions. Under the agreement, Microsoft is prohibited from employing student or teacher data for AI model training, tracking student activity, or allowing AI to make high-risk decisions without human oversight.
Known as the National AI Safety & Privacy Standard, this agreement was unveiled on September 9 by Microsoft Vice Chair and President Brad Smith, alongside Randi Weingarten of the American Federation of Teachers (AFT) and Michael Mulgrew of the United Federation of Teachers (UFT). School districts can incorporate these terms directly into their contracts with Microsoft, and non-compliance could lead to the termination of contracts and potential damages.
The agreement outlines ten enforceable protections, including a ban on the sale or ad use of student data, limited data collection, plain language explanations of tool functionality, 72-hour breach notifications, prompt security fix implementation, and annual certification and audits. Microsoft's Brad Smith emphasized that the agreement's core principles are privacy, safety, and transparency, and it will apply to all Microsoft districts starting November 1.
Teachers unions are pushing for similar AI privacy deals with OpenAI and Anthropic, given the lack of national consensus and the inadequacy of existing laws like COPPA and FERPA when it comes to generative AI.
Written by urgent.news from Times of India's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
