Major US teachers’ unions and Microsoft have announced the ‘National AI Safety & Privacy Standard’ to establish legally enforceable protections for children’s privacy and data, give schools real control over AI tools in the classroom, and provide parents with greater transparency and access to information.
The agreement was announced at a press conference on September 9 by American Federation of Teachers (AFT) President Randi Weingarten, United Federation of Teachers (UFT) President Michael Mulgrew and Microsoft Vice Chair and President Brad Smith.
Designed to protect students, educators and families, it allows school districts across the US to incorporate these protections into their agreements with Microsoft and hold the company accountable to them.
Today, I stood with the American Federation of Teachers President Randi Weingarten to announce the National AI Safety & Privacy Standard for Schools.
As outlined in our new agreement: “Children come first. Students are not products. Teachers are not beta testers. And schools… pic.twitter.com/WfAdnx24AT
— Brad Smith (@BradSmi) September 9, 2026
Microsoft’s US School Rollout
Beginning November 1, 2026, Microsoft will make the protections available to every US school district. The districts will be able to add the protections to new or existing agreements with Microsoft without waiting for a contract renewal or renegotiation.
The commitments will be contractually enforceable if Microsoft fails to meet them.
Microsoft is currently the only company to have signed onto the National AI Safety & Privacy Standard. AFT revealed that the union is in discussions with other technology companies, including OpenAI and Anthropic, about adopting the standard.
Weingarten said she hopes other AI providers will join the initiative, with the broader goal of establishing the framework as an industry standard.
“They’ve all been strong and willing partners of the Academy and I hope, soon, they will sign this, too,” Weingarten said. “And we have in the contract many, many, many, many different pages for other AI providers to sign because we want it to be an industry standard.”
The National AI Safety & Privacy Standard builds on the National Academy for AI Instruction, launched by AFT in July 2025 alongside the UFT, lead partner Microsoft, founding partner OpenAI and Anthropic. The $23 million initiative was designed to provide free AI training and curriculum to AFT’s 1.8 million members, beginning with K-12 (from kindergarten to twelfth grade) educators.
10 Protections for AI Use in Schools
The AI provider must comply with all 10 principles across every AI Provider Educational Product it provides to an EDU Customer that has opted in to the principles.
The agreement establishes 10 contractually enforceable protections covering AI training restrictions, data minimization, customer data ownership, human oversight and explainability, accountability, security, family consent and transparency, equity and accessibility, protection against feature creep and vendor lock-in, and long-term data responsibility.
“It starts to level the playing field in a competitive, ever-changing arena that lacks the guardrails our children and school communities need,” said UFT President Mulgrew, who leads the 200,000-member New York City AFT affiliate.
The bans on using data for AI training and selling covered data are permanent, while deletion and breach-related obligations continue after the contract ends.
“HIPAA (Health Insurance Portability and Accountability Act) and FERPA (Family Educational Rights and Privacy Act) never envisioned the advent of AI,” Weingarten said, “and in the absence of those protections, tech companies need to take responsibility for the products they create and market to students, educators and communities.”
She added, “And with New York City adopting our framework to ban screens and AI use in the younger years, this agreement adds to the momentum around online safety for kids and educators.”
Providers must maintain specified certifications, report breaches within 72 hours and address violations, with contractual remedies that can include pausing or ending AI use when required protections are not met.
The standard also requires providers to give schools a secure administrative way to access information about how their data is handled, including the types of data processed, retention periods, storage locations, sub processors and access controls.
Microsoft’s AI Safety Qualifications
The agreement with US teachers’ unions is effective for two years and is renewable.
Microsoft has not yet completed an independent ISO 42001 assessment for its AI Provider Educational Products and is targeting completion by December 31, 2027.
The restrictions do not apply to Microsoft’s Speaker Coach and Speaker Progress features when students initiate speaking activities, as the data is used only for coaching, feedback and progress insights.
Educators Facing Mounting Pressure
The announcement comes amid a broader backlash against the role of AI in schools, concerns among educators, growing scrutiny of Big Tech and a worldwide push to limit classroom screen time for younger children.
In a September 5 video message, Randi Weingarten said educators expressed deep disappointment over a lack of support, citing school funding cuts, AI, pressure to privatize public schools, book bans and restrictions on what teachers can teach and students can learn.
An AFT survey of more than 2,000 K-12 members conducted in August highlights the pressures facing educators, with nearly two-thirds reporting weekly stress levels of 7 or higher on a 10-point scale.
It found that 31% said they were likely to leave teaching within the next year, while only 36% said they would recommend their job to their own children.
AFT President Randi Weingarten described the situation as a “grave crisis” in the profession.
Also Read: New York City Mayor Zohran Mamdani Announces Ban on Gen AI in Schools
