Microsoft, OpenAI, and Anthropic are moving deeper into education through a new AI training effort for teachers. The National Academy for AI Instruction will open later this year in New York City, with an initial focus on kindergarten up to 12th grade instructors in the American Federation of Teachers.
The academy is framed as a way to help educators understand and use AI tools in classrooms, rather than leave schools reacting to them after students have already adopted them. It also raises a larger question: who should shape the rules, standards, and habits around AI in education?
A Training Center Built Around Classroom AI
The National Academy for AI Instruction is described by the union as a first-of-its-kind $23 million initiative funded by tech companies. Microsoft and OpenAI announced on Tuesday that they are helping launch the center, and Anthropic, which develops the Claude chatbot, recently became a collaborator.
The first audience will be AFT members teaching kindergarten up to 12th grade. Over time, the AI training will be open to all members of the federation, which says it represents about 1.8 million workers, including K-12 teachers, school nurses, and college staff.
The academy and its partners are seeking to support about 400,000 union members over the next five years. The source article describes that as about 10 percent of all teachers nationwide.
The program will include workshops and online courses. According to the press release cited in the source article, the curriculum is designed by “leading AI experts and experienced educators” and will count for continuing education credits. The academy will be operated “under the leadership of the AFT and a coalition of public and private stakeholders,” the release added.
Why Schools Are Looking for Guidance
Schools have spent the past few years trying to keep up with student use of AI chatbots, including OpenAI’s ChatGPT, Microsoft’s Copilot, and Google’s Gemini. These systems can help write papers and solve some math problems, but the source article notes that they can also produce confident and costly errors.
That combination has created a complicated classroom problem. AI tools can support learning, planning, and creativity, but they can also blur the line between help and substitution. Parents, educators, and employers have voiced concern about whether chatbots prevent students from building essential skills themselves.
Some school districts have responded with tools meant to catch AI-assisted cheating. Teachers have also begun building lessons around responsible use of generative AI. At the same time, educators are using AI to help with time-consuming tasks such as developing teaching plans and materials.
The case for training is straightforward: if AI tools are already present in schoolwork, teachers need practical knowledge of how they work, where they fail, and how to set expectations. Without that, classrooms risk being governed by uneven rules, inconsistent enforcement, and trial-and-error policies.
The Union Wants a Seat at the Table
AFT president Randi Weingarten has argued that educators should be involved in decisions about how AI enters their profession. During a press conference on Tuesday, she said, “Teachers are facing huge challenges, which include navigating AI wisely, ethically, and safely.”
She also connected the issue to the arrival of ChatGPT. “When we saw ChatGPT in November 2022, we knew it would fundamentally change our world. The question was whether we would be chasing it or we would try to harness it.”
That framing explains why the academy matters beyond a single training program. The center could help teachers understand fast-changing AI technologies and adapt curriculum for a world in which these tools are core to many jobs.
OpenAI’s chief global affairs officer, Chris Lehane, made a similar argument about access to AI’s benefits. He said on Tuesday that the spread of AI and related productivity gains were inevitable. “Can we ensure those productivity gains are democratized?” he said. “There is no better place to begin that work than the classroom.”
The Commercial Tension Is Hard to Ignore
The initiative is also likely to face criticism from some union members who are wary of technology companies influencing what happens in US classrooms. The source article notes that Google, Apple, and Microsoft have competed for years to get their tools into schools in hopes of turning children into lifelong users.
That concern does not disappear because the program is framed as teacher training. If the tools, curriculum, or standards are shaped partly by companies that sell AI products, educators may ask whether classroom needs or commercial incentives are setting the agenda.
The tension is wider than the United States. Just last week, several professors in the Netherlands published an open letter calling for local universities to reconsider financial relationships with AI companies and ban AI use in the classroom.
Still, the source article suggests that all-out bans appear unlikely as generative AI chatbot use grows. That leaves AI companies, employers, and labor unions trying to find common ground, even when their interests do not fully match.
What Remains Unclear
The academy follows a Microsoft partnership announced in December 2023 with the American Federation of Labor and Congress of Industrial Organizations on developing and deploying AI systems. The American Federation of Teachers is part of the AFL-CIO, and Microsoft had said at the time that it would work with the union to explore AI education for workers and students.
One major unresolved issue is how the new training will fit with local policies on AI use. Those policies are often set by elected school boards, and the source article says it is unclear how the academy’s work will intersect with them.
That question matters because teacher training alone cannot settle every classroom rule. Schools still need decisions on when AI help is acceptable, how assignments should change, what counts as cheating, and how students should learn to evaluate chatbot outputs.
For now, the National Academy for AI Instruction represents a significant attempt to move teacher AI training from informal experimentation into a larger institutional structure. Its success will depend not only on workshops and online courses, but on whether educators trust the process and whether the training helps them make clearer decisions in real classrooms.