Technology news and discussion on TalkFuse.
Teachers Union Puts Anthropic On A Leash Over Student Data
Anthropic has signed a memorandum of agreement with the American Federation of Teachers and the United Federation of Teachers to set privacy and safety requirements for AI used in schools. The agreement matters because it seeks to protect pupils’ data and keep educational decisions under teachers’ control, amid growing concern about AI’s role in classrooms and limited federal regulation.
The unions say the agreement legally bars the sale of student data and requires clear information about AI products to be shared with schools and parents. AFT president Randi Weingarten has previously called for restrictions on screen use and student-facing AI for younger pupils, as well as chatbot companions for under-16s. Commentators differed: one called for classroom AI to be banned, while parent advocate Nicole Runyon described the agreement as a positive step and stressed the risks of children sharing personal feelings with chatbots.
In short
- Anthropic agreed to school AI privacy and safety requirements.
- Student data cannot be sold under the agreement.
- The deal drew both support and calls for stricter limits.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
The agreement could give schools enforceable safeguards against selling pupils’ data and require clearer information for families, addressing risks that may outpace federal regulation. Supporters can also argue that setting boundaries while keeping teachers in control allows schools to assess useful AI tools without leaving children’s privacy and educational decisions to companies.
The case against
Critics may argue that an agreement with unions cannot substitute for comprehensive, consistent regulation or guarantee how AI systems will affect pupils in practice. Given concerns about children sharing sensitive information with chatbots and the effects of screen use, they may favour restricting or banning student-facing AI, particularly for younger pupils, rather than relying on safeguards around its use.
Explained simply
Anthropic is a US company that develops artificial intelligence tools, including chatbots that can generate text and answer questions. Schools are increasingly exploring these tools for teaching and administration, raising questions about what information pupils share and how AI may affect learning.
The American Federation of Teachers and the United Federation of Teachers are US unions representing educators. Their agreement with Anthropic sets privacy and safety requirements for the company’s school AI products, including limits on using pupils’ data and requirements to explain products to schools and parents.
The issue matters because children may share personal information with AI systems, while schools and families need to understand how those systems are used. In the United States, federal rules specifically governing AI in education remain limited, so agreements between companies and unions are one way of setting expectations.