Major teachers unions and education associations are working to incorporate diversity, equity and inclusion principles into artificial intelligence tools as schools prepare to adopt the technology more widely, according to a report released last month by the watchdog group Defending Education.

The July 22 investigation argues that organizations including the National Education Association and American Federation of Teachers are emphasizing “algorithmic bias” and equity in guidance, model policies and training programs. Defending Education contends this approach risks institutionalizing politicized content in classroom AI systems once the tools become standard for lesson planning, student research and feedback.

“The integration of AI into education is not a matter of if but when,” the report states. It says unions and associations aim to shape the technology first so that DEI considerations become built-in features of systems used in K-12 schools.

The National Education Association, which says it represents about 3 million members, maintains an “AI in Education Hub” that the union describes as “grounded in equity, ethics, and educator voice.” The NEA offers courses such as “When Technology Isn’t Neutral: Understanding AI and Bias” and “Promoting AI Equity and Access.”

The union has also published a sample school board policy that districts can adopt. The policy calls for regular audits of AI tools “to identify and mitigate biases,” the creation of oversight committees to ensure alignment with “district equity goals,” and a requirement that vendors meet standards for “transparency, equity, and ethical decision-making.”

In a section on algorithmic bias, the policy defines the term as “systemic, unwanted unfairness in how a computer detects patterns or automates decisions” often based on characteristics including race, gender, ethnicity and sexuality. A 2024 NEA task force report on artificial intelligence in education stressed the need for attention to whether AI systems “further perpetuate racial injustice and social inequities.”

The American Federation of Teachers, representing about 1.8 million members, issued 2024 guidance titled “Commonsense Guardrails for Using Advanced Technology in Schools.” One guardrail states that technologies “must be deployed so that they advance equity and fairness in public education and cannot be adopted so that they widen, rather than help close, the digital divide and other inequities.”

The AFT has launched the National Academy for AI Instruction, a $23 million initiative with the United Federation of Teachers, Microsoft, OpenAI and Anthropic. The program is intended to provide free AI training and curriculum to educators. A 2025 AFT resolution on the academy noted that without proper guardrails, AI “can perpetuate bias, invade privacy, and undermine educational equity.”

Other groups cited in the report include the California Teachers Association, which helped develop state guidance urging educators “to approach AI with an equity lens”; the American Association of School Administrators, which the report says has referenced concerns about AI development by “white males” and the concept of a “New Jim Code”; and the National Association of Elementary School Principals and National Association of Secondary School Principals, both of which have published materials recommending that schools question vendors about algorithmic biases and ensure AI tools advance equity.

Algorithmic bias claims as often tied to the idea that AI systems produce answers reflecting those creators’ worldviews. The group argues most AI systems are designed to deliver accurate responses based on available data rather than identity-based adjustments.

The report focuses on the largest unions and associations and notes it is not a comprehensive survey of all education organizations. It does not claim the groups oppose AI use outright, but rather that they are seeking to shape its development and adoption around equity frameworks.

Teachers unions and the partnered technology companies have publicly framed their efforts as ensuring responsible, ethical and effective classroom use of AI while protecting student privacy and educator judgment. The National Academy for AI Instruction, for example, describes its mission as meeting challenges so that AI is “safe, fair, effective, and ethical.”

Defending Education’s investigation arrives as school districts across the country evaluate AI tools and as states and professional groups develop related policies and training.