Microsoft has new AI privacy rules for schools

Amid a policy crackdown, it’s agreeing to avoid training on student data and prohibit AI companions.


Microsoft agreed to a set of safety and privacy principles for AI in schools a week after two major school systems announced a ban on student-facing AI.
In a new agreement with the American Federation of Teachers (AFT), the second-largest teachers union in the US, and its New York City affiliate the United Federation of Teachers (UFT), Microsoft committed to ten principles that can be contractually enforced by school districts that adopt them. The terms include pledging not to train AI models on student or educator data, limiting the amount of data Microsoft collects in the first place and disclosing to families how its tools work in plain language, prohibiting AI companions, and requiring human review for “high-risk” decisions.
The agreement comes after New York City and Los Angeles both implemented one-year bans on AI tools for many students, giving the districts time to consider the proper use and guardrails. As parents and some educators have resisted the use of AI and screens more generally in schools, Microsoft likely recognizes that similar decisions could be coming. Beginning in November, school districts can opt to add the terms to new or existing agreements without having to renegotiate their entire contracts.
“[N]o one else, including the federal government, has stepped up to do the real work”
The AFT, for its part, has been critical of tech in schools, but also open to working with tech companies to prepare for the current AI wave. In a May speech , AFT President Randi Weingarten called for a ban on screens before third grade as well as “student-facing AI” before middle school. But the group also announced a $23 million AI training hub for educators funded by Anthropic, Microsoft, and OpenAI.
In a statement about the new privacy agreement with Microsoft, Weingarten said the terms were “iron-clad” and necessary “because no one else, including the federal government, has stepped up to do the real work.” She added, “We can get angrier and angrier, or we can act decisively; anything less than legally enforceable provisions is simply a wish list.”
Verified source · The Verge
Reported by The Verge. Open the original for full media and formatting.
More in Policy
All news
PolicySuperintelligence is coming. Should we let it?
AI companies have been talking about superintelligent AI like it’s inevitable, but recent safety incidents like OpenAI’s Hugging Face breach are demonstrating the potential dangers of deploying AI systems that are more capable than humans. So what happens when we can’t reliably…
Read at TechCrunch
PolicyControlAI’s Connor Leahy on why superintelligence is ‘not a weapon, it’s an adversary’
AI companies have been talking about superintelligent AI like it’s inevitable, but recent safety incidents like OpenAI’s Hugging Face breach are demonstrating the potential dangers of deploying AI systems that are more capable than humans. So what happens when we can’t reliably…
Read at TechCrunch
PolicySuno replaces its AI models with a new one trained on licensed music as copyright suits pile up
As it grapples with a bevy of lawsuits, Suno said its new model, Suno v6, is not trained using music it used to train previous versions of the AI model.
Read at TechCrunch
PolicyMore than 1 in 10 chance AI ‘could kill all humans,’ says Anthropic safety lead after colleague quits
A senior Anthropic safety researcher has said there is more than a 10 percent chance artificial intelligence "could kill all humans" by the end of the decade, just hours after a colleague resigned over fears the AI lab and its rivals are carelessly racing to build "superhuman sy…
Read at The Verge