The public is waking up to the dangers of AI: In the past month alone, activists in California, Texas, Oregon, and other states have tried to stop data centers and their voracious toll on water and energy.
However, most people remain largely unaware of another kind of damage wrought by AI, in concert with DEI rollbacks — damage with deep consequences for American STEM competitiveness, and, more importantly, fairness and equality: We are squandering our investments in Black STEM students.
For three decades, the United States made a deliberate bet. Government agencies, universities, and private corporations invested billions of dollars in programs designed to bring more Black students into STEM fields. This was a critical strategy to improve global competitiveness and innovation by drawing on all of America’s best minds. The National Science Foundation (NSF), for example, funded research and pipeline programs to attract and retain underrepresented students in STEM.
Now, Black STEM undergraduates and recent graduates across the country are being squeezed from two directions at once: the rapid rollback of DEI programs that helped them access, persist in, and launch careers in STEM, and the simultaneous rollout of artificial intelligence that is eliminating the entry-level jobs they were trained to fill.
In the first quarter of 2025 alone, the NSF terminated dozens of DEI-focused grants, leaving researchers unable to continue their work and students scrambling for support. At the same time, major corporations such as Google, Meta, Ford, and Target quietly dismantled their racial equity commitments, once seen as assets and now treated as liabilities as the policy landscape shifted.
This isn’t just policy change. It is abandonment — and it is happening in the middle of a technological revolution that demands more STEM talent, not less.
At the very moment DEI support structures are being dismantled, AI is reshaping the STEM labor market in ways that disproportionately harm Black students. Research from McKinsey and Brookings shows that automation affects Black workers at higher rates, because Black workers are disproportionately concentrated in the entry-level and mid-skill roles most vulnerable to displacement.
Entry-level engineering and technical roles — typical on-ramps to STEM careers — are among the first being automated away. For Black students, who often lack legacy networks, family connections, and generational access, those entry-level jobs were not just jobs. They were the gateway to brighter futures. Remove the gateway, and the path disappears.
When we invest in the students who have faced the most systemic barriers (students who have been marginalized for more than 400 years), we create conditions that lift everyone.
The research is clear. Programs designed to support students who face the greatest obstacles tend to improve outcomes across the board. For instance, Universal Design for Learning, a framework originally developed to serve students with disabilities, is now considered best practice for all learners. Likewise, HBCUs produce a disproportionate share of Black STEM undergraduates who go on to earn doctorate degrees. A 2025 study examining identity affirming and culturally responsive practices concluded that these practices were beneficial for all students.
Further, research shows that diversity supports innovation. A 2022 study published in the Proceedings of the National Academy of Sciences (PNAS) found that diverse scientific teams produce more innovative and more highly cited research than homogenous teams. The National Academies of Sciences found that diversity in STEM workforces drives economic productivity and problem-solving capacity. America’s global competitors — China, the European Union, India — are not pulling back from STEM talent development, but we are.
By dismantling DEI structures in the age of AI, we are choosing, in this moment, to squander our best investments. To stop this, we need to promote the success of Black STEM students in at least three ways.
First, vote for people and policies that support inclusion in STEM education. Legislators set the conditions under which universities, federal agencies, and funding bodies operate. Second, companies that have rolled back their DEI programs must reinvest — not through checkbox diversity initiatives, but through programs that genuinely reach all workers. Third, legislators must do the serious work of data-driven policymaking. Before cutting programs, study the outcomes. Before declaring DEI initiatives ineffective, examine the evidence.
The rollback of DEI and the rollout of AI are not separate crises. Together, they form a double blow that is landing hardest on the students we have spent decades trying to include. If we allow this moment to squander that investment, we will not only have failed those students. We will have failed ourselves.
Roni Ellington, PhD, is a professor of mathematics education at Morgan State University, an HBCU, and a Public Voices Fellow on Youth Well-Being and Power of the OpEd Project and Hopelab.
