AFT accepts $23M for AI training

The American Federation of Teachers did not newly accept the money this week. It announced the $23 million initiative on July 8, 2025, when it launched the National Academy for AI Instruction with Microsoft, OpenAI, Anthropic and the United Federation of Teachers. The program was pitched as a free AI training and curriculum effort for AFT members, beginning with K-12 educators, and tied to a physical training hub in Manhattan. The recent attention came after renewed coverage, including a Guardian report published August 5, 2026, revisited the arrangement and the criticism around it. The core facts, though, trace to the 2025 launch: AFT said the academy would serve its 1.8 million members, while outside coverage said the effort targeted 400,000 teachers for training by 2030. (aft.org) ### Who is paying for the training? Microsoft, OpenAI and Anthropic were named as the corporate backers when the academy was announced. AFT described Microsoft as the lead partner, OpenAI as a founding partner and Anthropic as a partner in the initiative. The total funding announced was $23 million. The union’s own materials say the academy is designed as an educator-led national training hub focused on AI use in public education. (aft.org) AFT’s academy site says the project is meant to shape AI use in schools around safety and “people-first technology.” ### What is the academy supposed to do? AFT said the academy would provide free AI training and curriculum for its members, starting with K-12 educators. The Manhattan site was described as a bricks-and-mortar facility where teachers could learn how to use AI in classroom work and school tasks. (aft.org) Outside reporting gave more detail on the classroom uses being contemplated. (aiinstruction.aft.org) Time reported that workshops would cover tasks such as generating lesson plans, quizzes and emails to parents. K-12 Dive reported that the initiative’s target was to help teachers build “AI fluency skills” over a five-year period. ### Why has the partnership drawn criticism? The criticism centers on the source of the money and the role of technology companies in public education. (aft.org) The Guardian said critics raised concerns about conflicts of interest and about large AI companies gaining influence over how teachers are trained to use products that those same companies build. (time.com) AFT and allied union officials have framed the partnership differently. UFT President Michael Mulgrew said the union did not want technology to be “weaponized” against educators, and AFT’s formal resolution on the academy says educators must remain at the center of decisions about how AI is used in schools. ### Is this a new policy shift by the union? AFT’s own documents suggest the academy fits into a broader union strategy rather than a sudden change. (catherwood.library.cornell.edu) A 2026 AFT resolution says the union had already been pushing for “responsible, ethical, and safe” use of AI in schools and had argued for guardrails on privacy, bias and educator autonomy. (uft.org) That matters because the union is presenting the academy as both training infrastructure and a way to shape adoption. The academy website says it is “designed by educators,” while the launch announcement said the goal was to put educators “in the driver’s seat” as AI enters classrooms. ### What happens next? The next concrete milestone was the academy’s planned opening in Manhattan in fall 2025, according to the launch announcement and subsequent coverage. (defendinged.org) AFT said the training would begin with K-12 educators, and outside reports said the longer-run goal was to reach 400,000 teachers by 2030. (aft.org) (aiinstruction.aft.org)