Exclusive: Virginia School District Injects DEI Into AI Tools

A school district in Virginia is using a diversity, equity, and inclusion (DEI) filter on its use of generative AI (GenAI) in schools to inject a race-focused lens into teacher and administrative decision-making.

According to documents obtained by Defending Education, Albemarle County Public Schools (ACPS), a far-left school district surrounding Charlottesville, and which includes Thomas Jefferson’s Monticello, has built its GenAI policy based around its “anti-racism” policy, which is inherently discriminatory against white students in favor of other races.

“While schools should engage in good judgement and do their due diligence when it comes to Artificial Intelligence integration, the fact that the district is vetting AI based on its compliance with diversity, equity, and inclusion should be concerning for parents,” Rhyen Staley, Director of Research at Defending Education, told The Federalist. “By only allowing the use of AI and information sources that reflect a leftwing political bias, district administrators are setting a precedent that is harmful to the learning process and neutrality of schools.”

AI is set to be used by both students and teachers.

A May 21, 2025, email from ACPS Director of Equity Education Ayanna Mitchell details a “GenAI Equity Prompt Sheet” to be used as a “tool” when “making decisions about GenAI integration.”

Mitchell shared a “draft tool designed to support subcommittees in making equity-centered decisions throughout their work. It’s intended to surface questions around access, bias,
stakeholder voice, and policy alignment in a practical, usable format.”

Chief among these considerations is the question, “Does this tool or policy reflect the values in ACC, IGAK, and JBA? Anti-racism, cultural responsiveness, and gender inclusivity? Student agency, safety, and identity affirmation?”

IGAK refers to the school district’s Equity Education Policy, under which house “ACC,” meaning “anti-racism,” and “JBA,” meaning “Policy on the Treatment of Transgender and Gender Expansive Students.”

As The Federalist reported, ACPS already forces its students to adhere to radical ideologies, including DEI and anti-racism, as part of “core competencies” to be achieved before graduation.

ACPS’s AI policy reads similarly, stating under its “guiding principles” that one of the “challenges” it faces is “cultural erasure.”

“AI tools may default to dominant cultural norms or fail to recognize underrepresented or marginalized groups and aspects of their culture, creating risks of invisibility and harm,” it states.

Ironically, many AI tools have been found to do the precise opposite, actually defaulting to non-white persons to represent historical figures like America’s Founding Fathers, as was infamously the case with Google’s Gemini, which also created “diversity Nazis,” who were Asian and black.

Defaulting against white persons is built into most AI tools because it dredges and learns from the same kinds of “anti-racism” ideology that ACPS employs. AI models see the Founding Fathers as not diverse enough, so it responds by depicting them as historically inaccurate races in an effort to, ostensibly, not be racist.

When asked about this concern, ACPS chief communications officer Jason Grant refused to answer, instead pretending he is unaware of the overwhelming racial problems posed both by ACPS’s race-centric policies, and those policies being interwoven into an already DEI-oriented AI paradigm.

Even a study attached to Mitchell’s email admits, “When given explicit race related cues, ChatGPT actually, contrary to an expectation of bias, scored Black students higher.”

That article does not consider automatically scoring black students higher as an inherent “bias” because under the “anti-racist” ideological framework, “bias” only exists if it favors white people. It decidedly does not if it favors non-white people over white people.

Other considerations in ACPS’s AI tool include asking whether the “needs of historically marginalized students explicitly considered?” or “How will this affect students with disabilities, multilingual learners, LGBTQ+ youth, or those in high-poverty areas?”

Another asks, “Have community or affinity groups raised concerns we haven’t yet addressed?,” referencing explicitly race- or gender-segregated “affinity groups.” One prompt also asks “What historical inequities might this unintentionally reinforce?”

The prompts in the email appear to vaguely reference “equity grading” and the de-emphasis on addressing behavioral issues, which are part of ACPS’s broader DEI framework, by asking “What assumptions are built into this technology or policy? Who is assumed to be the ‘average’ learner? Are cultural norms, language use, or behavior expectations embedded in ways that might exclude some students?”

As The Federalist reported, “Ignoring behavioral and attendance issues, otherwise known as restorative justice, is often a system that works to hold some students of some races or ethnicities to different standards than others.”

“These prompts are not intended to slow innovation, but to anchor it in justice, empathy, and accountability,” the email states.

Mitchell’s email also included “optional readings that informed the development of this prompt sheet,” shedding light behind the thinking of ACPS in its development of the use of AI tools and policies.

One called Bringing Communities In, Achieving AI for All, considers that broader societal implications about AI’s impact on people’s “livelihoods” are considerably less important. “More concerning,” the article states, “is the mounting evidence showing that the output of AI models exacerbates social inequity and injustice.”

Far from being concerned about the provable anti-white bias in many AI models, the article considers “those with the most at stake” to be “marginalized people,” which explicitly excludes white persons.

However, perhaps the most telling portion of that article is the groups it recommends for help in “focus[ing] the attention of policymakers and regulators” toward a more DEI-focused approach to AI.

The Ford Foundation, which it describes as “exemplary in this regard,” is perhaps most famous for funding anti-Immigration and Customs Enforcement (ICE) radicals and other terrorists, rigging elections, and bankrolling a host of other radical-left agitation movements across the country.

As the article states, the Ford Foundation also funds tech company Fight for the Future, which is also recommended by the article to advance DEI integration with AI.

As The Federalist reported, Fight for the Future advances gender ideology, claiming men can be women, and also opposed the “Kids Online Safety Act,” which would have required online platforms to take “reasonable measures” to protect children from “sexual exploitation and abuse.”

Perhaps that is no surprise, because aside from taking Ford Foundation, Fight for the Future also took money from Aylo, the parent company of PornHub, which has profited from sex trafficking.

Breccan F. Thies is the White House correspondent for The Federalist. He is a co-recipient of the 2025 Dao Prize for Excellence in Investigative Journalism. As an investigative journalist, he previously covered education and culture issues for the Washington Examiner and Breitbart News. He holds a degree from the University of Virginia and is a 2022 Claremont Institute Publius Fellow. You can follow him on X: @BreccanFThies.