Microsoft, the American Federation of Teachers (AFT), and the United Federation of Teachers (UFT) have announced a first-of-its-kind National AI Safety & Privacy Standard designed to establish legally enforceable protections for students, families, and educators in U.S. schools.
The new standard arrives at a moment when artificial intelligence is increasingly embedded in education technology, and concerns about data privacy, surveillance, and algorithmic bias in schools have never been higher. This agreement offers a concrete framework that schools and districts can incorporate directly into their procurement agreements with Microsoft.
What the Standard Guarantees
According to the announcement, the National AI Safety & Privacy Standard includes four key protections:
- No AI training on student data: Tech companies are prohibited from using student information to train AI models.
- No student tracking: Students cannot be tracked or subjected to behavioural profiling through their use of school technology.
- Human oversight of AI decisions: AI cannot make consequential decisions about students or educators without meaningful human involvement.
- Transparency about tools: Companies must be transparent about how AI tools function and what data they collect.
The standard is legally enforceable because U.S. school districts can incorporate these protections directly into their customer agreements with Microsoft. This moves the conversation from voluntary pledges to binding contract terms.
Why This Matters for Schools
Education has long been a testing ground for new technology, but it has also been a sector where data protection has lagged behind adoption. High-profile cases—such as Austria's finding that Microsoft tracked students—have raised alarm among parents, teachers, and regulators about the trade-offs schools face when choosing tools.
For schools weighing the adoption of AI-assisted tools in platforms like Microsoft Teams, this standard offers a clearer basis for evaluating whether those tools meet local and national privacy expectations. It also sets a precedent: if a major vendor can commit to these protections, smaller competitors and new entrants should be expected to follow suit.
A Teacher-First Lens
The involvement of the AFT and UFT is significant. Both unions represent hundreds of thousands of educators who are often asked to integrate new technology into their classrooms without clear guidance on privacy implications. By co-authoring this standard, they have moved teacher voice from the sidelines to the centre of the conversation about AI in schools.
This matters because teachers are not just users of technology—they are stewards of student data and gatekeepers of classroom trust. A standard that carries union backing is more likely to be understood, adopted, and enforced at the classroom level.
What It Means for Ed-Tech Providers
For providers building tools for schools, this announcement raises the bar. Compliance with GDPR and other data-protection frameworks has long been a baseline expectation in Europe, but the U.S. market has been more fragmented. This new standard may prompt more uniform expectations around AI transparency and data use across American school districts.
At Simpleclass, we've designed our platform with these principles already in mind. We host all video, recordings, files, and chat in the EU, and we do not use student data to train algorithms or track behaviour. Our access model is built on account and course enrolment, not shareable meeting links, which reduces the risk of data leaks or unauthorised access. We believe privacy should be a feature of the architecture, not a legal afterthought.
Looking Ahead
It remains to be seen how widely this standard will be adopted beyond Microsoft's customer base, and whether other major platforms will commit to similar terms. The announcement does not yet detail enforcement mechanisms or what happens if a company breaches the standard.
Still, this is a meaningful step. It acknowledges that AI in education is not neutral, that data practices have consequences, and that teachers, students, and families deserve a say in how those tools are built and deployed.
As schools continue to navigate the promise and peril of AI, frameworks like this one offer a template for holding technology accountable to the people it is meant to serve.