Over the past decade or so, diversity, equity, and inclusion initiatives have transformed America beyond recognition, effectively replacing performance-based standards with identity-based quotas. Workplace hires, college admissions, and corporate promotions swapped personal achievement for group membership. Thankfully, commentators like Piers Morgan say those days are behind us. Woke, we’re told, is dead and buried.
It isn’t. DEI not only refuses to die but appears stronger than ever.
Administrators are writing their preferred political rules into machine intelligence.
A recent report by Defending Education, a grassroots organization dedicated to protecting American children from harmful ideologies, shows how universities are embedding DEI ideas directly into the AI systems that will shape how Americans learn and work.
Perhaps the most troubling part of this report is how ordinary the process can look. DEI can be built into the lessons used to teach children what AI is, the tests used to judge whether AI is “fair,” and the rules universities recommend for how AI should behave. The old social contract rewarded individual effort, clear competence, and the ability to tell fact from fiction. The new digital contract ranks software output based on ideological alignment.
Behind the academic curtain
MIT, for example, is developing AI lessons for K-12 students that teach kids to think about “algorithmic bias” and identify when an AI system produces unequal results. An algorithm that flags a high-risk loan applicant based on credit history fails this test if certain demographic ratios look uneven. UC Berkeley goes further, describing AI as something that can reinforce racial, gender, and other inequalities. Its researchers are developing tools to measure whether AI produces supposedly “equitable” results for different groups. Computer programmers used to debug code to fix technical errors. Today, researchers rewrite code to ensure equal outcomes across groups.
Columbia is developing tests for whether chatbots work properly for LGBTQ communities and has studied how prompts can produce better responses for “queer BIPOC youth.” A mathematical tool designed to process human language is being reprogrammed to act like a pink-haired guidance counselor with a radically leftist agenda. Other universities run programs that bring ideas about “equity” and AI directly to K-12 students and teachers and publish research that examines AI through Marxist ideas.
RELATED: Parents weren’t asked. Many still don’t know. Inside the new school surveillance nightmare.
gorodenkoff/Getty Images
A snowball effect is forming. Universities are helping write the rules for how AI will be taught and judged.
A century ago, political activists seized physical infrastructure like telegraph offices and train stations to control public life. Modern ideologues target software compilers and language models instead. Control the training data, and you control every answer the machine gives to the public.
Automating indoctrination
That creates a pipeline: First, a university develops a curriculum. Teachers receive it. Students learn from it. Then, AI companies adopt these standards to avoid lawsuits. Policymakers use university research when writing regulatory rules. Before long, ideas that began inside academia become part of the technology tens of millions of Americans use every day.
Engineers built computers to give accurate answers based on hard data. The current academic push forces these machines to censor objective reality whenever uncomfortable truths conflict with social theories. Ask a neutered AI model simple historical questions, and it will distort facts to avoid offending academic sensibilities. Ask it to evaluate job applications, and it will prioritize demographic targets over verifiable skills. Ask it about the perpetrators of crime in certain cities across America, and it will do everything but provide accurate information. White readers, particularly straight white men, routinely described as society’s most “privileged” group, should be especially concerned.
DEI on its own is dangerous enough, but merged with arguably the most powerful technology ever created, it is a recipe for disaster.
The end result is AI that lies to its users by design. When air traffic control systems, medical diagnostic tools, and financial software rely on algorithms trained to prioritize equity over accuracy, real-world performance drops and real lives are at risk.
Enforced mediocrity
The original social contract built American prosperity on a simple promise: Work hard, master a trade, deliver superior results, and you climb the ladder. The bureaucratic class spent decades dismantling that promise inside taxpayer-funded institutions. Now, those same administrators are writing their preferred political rules into machine intelligence. Human bias can be challenged, debated, and defeated in public debate. Software code hidden inside a proprietary algorithm operates without public oversight.
If left unchecked, this rather diabolical system will embed identity-based quotas into the permanent digital infrastructure of everyday life. Citizens will interact with technology that enforces destructive doctrines while presenting them as non-negotiable facts. The death of merit was bad enough when enforced by university deans and corporate HR departments. Machine-enforced mediocrity could make those bureaucrats look like absolute amateurs.
All is not lost, however. Now that the public can see what universities are doing, they can be challenged. The Trump administration has already ordered universities to abandon DEI. It now needs to recognize where DEI is taking root next and respond accordingly.
Artificial intelligence, Tech
