DOJ: 2 Years of Duke Law Admissions Data Reveal Racial Proxy Use, EdTech Scrambles

For edtech companies powering college admissions, the DOJ's investigation of Duke Law is a wake-up call. With two years of admissions data showing LSAT score gaps by race, the case exemplifies how even well-intentioned holistic review platforms can become vehicles for illegal discrimination. EdTech vendors must quickly retool algorithms, fine-tune essay prompts, and introduce bias-detection dashboards—or risk becoming the next target in a widening regulatory net.

The Department of Justice issued a findings letter to Duke University School of Law on August 6, 2026, alleging that the institution illegally considered race in its admissions process in violation of the Supreme Court's 2023 ruling banning race-conscious college admissions. DOJ officials assert that while Duke Law claimed to follow race-neutral policies, it used other factors as substitutes for race, notably personal essays and whether an applicant was a first-generation college student, to achieve a racial balance. The letter is the first major enforcement action under the landmark Students for Fair Admissions v. Harvard and UNC decision, signaling an aggressive new phase in the Trump administration's scrutiny of admissions practices.

“

The letter is the first major enforcement action under the landmark Students for Fair Admissions v.

The DOJ's investigation uncovered statistical evidence that Black or Hispanic applicants were more likely to be admitted than white or Asian applicants with the same academic credentials. For the 2024 and 2025 admissions cycles, the median Law School Admission Test (LSAT) scores of rejected white and Asian applicants exceeded the median scores of admitted Black applicants. This disparate impact, combined with the use of proxy variables, forms the core of the government's case that Duke Law engaged in intentional discrimination.

Assistant Attorney General Harmeet K. Dhillon of the Civil Rights Division starkly warned that using diversity essays to identify preferred races constitutes illegal discrimination, regardless of perceived good intentions. Her statement underscores that the DOJ will treat any mechanism that functions as a racial preference—even if superficially race-neutral—as a violation of civil rights law. Duke Law responded that it is reviewing the letter and remains committed to following the law "in a manner that is consistent with our academic mission," but offered no immediate change in policy.

The 2023 Supreme Court decision, decided 6-3 along ideological lines, eliminated the consideration of race in admissions for higher education institutions. However, the Court left a narrow opening for universities to consider applicants' discussions of how race affected their lives, provided such discussions are tied to qualities like character or determination. The DOJ's findings suggest that Duke Law exploited this exception by designing essay prompts and holistic review criteria that inevitably reveal applicants' race and then using that information to engineer outcomes. First-generation college status, often correlated with race, is now under scrutiny as a potential backdoor for racial balancing.

The implications for higher education are profound. If the DOJ's interpretation prevails, any admissions factor that closely correlates with race could be challenged as a proxy, including socioeconomic status, zip code, or even participation in certain extracurricular activities. Universities that have publicly committed to maintaining diversity will face a stark choice: risk federal investigation and potential loss of funding or abandon any criterion that yields disparate outcomes. The legal concept of "disparate impact"—where a neutral policy produces racially unbalanced results—may become a flashpoint, with the DOJ arguing that intent can be inferred from statistical disparities.

Market impact extends to the edtech sector. Admissions management platforms, AI-driven holistic review tools, and enrollment analytics companies must now enable clients to audit and strip out any variable that could be construed as a race proxy. Companies like Common App, Liaison International, Technolutions, and emerging AI admissions startups will face heightened demand for compliance verification features, but also potential liability if their algorithms unwittingly perpetuate bias. The legal tech sector will similarly see a surge in demand for audit tools and predictive models to assess risk under the new enforcement regime.