Trump's 'Preventing Woke AI' order threatens to embed racial bias in healthcare systems
President Donald Trump signed a sweeping executive order titled “Preventing Woke AI in the Federal Government Act,” an action that deliberately attempts to erase terms like diversity, equity, and inclusion (DEI) and roll back efforts to address systemic racism in federal artificial intelligence systems.
For experts in medicine and health equity, this is not just political posturing; it threatens lives by jeopardizing years of work to identify and correct structural biases that have long harmed marginalized communities, particularly Black Americans.
AI Bias and Medical Accuracy
AI is rapidly transforming healthcare, used to triage emergency room patients, prioritize follow-up care, and predict disease risk. However, these algorithms are trained on real-world data that is inherently biased.
A 2019 study in Science demonstrated this issue by examining a widely used commercial healthcare algorithm designed to flag high-risk patients. Researchers found the algorithm was silently using a proxy for clinical need: the amount of money previously spent on a patient’s care. Because Black patients typically receive less care for the same symptoms, this spending proxy caused the algorithm to drastically underestimate their need. While nearly 46.5% of Black patients should have been flagged for care, the algorithm identified only 17.7%.
This is not an isolated case. Racism has become embedded in code through methods such as:
-
Kidney Function (GFR): Equations for Glomerular Filtration Rate long included a "correction factor" for Black patients, based on unscientific assumptions. This adjustment inflated kidney scores, often delaying eligibility for transplants or specialty care.
-
Pulmonary Function Tests (PFTs): These tests, used to diagnose lung diseases, often apply a race-based correction assuming Black people naturally have lower lung capacity, contributing to underdiagnosis.
The Danger of the Executive Order
In recent years, researchers have pushed back, leading many hospitals to remove these race-based corrections. Trump’s executive order threatens to shut down this crucial work.
By banning federal agencies from considering systemic racism or equity in AI development, the order effectively outlaws the very efforts needed to fix these problems. Supporters claim the order promotes "neutrality," but in a system built on inequity, neutrality serves only to reinforce the existing biases it pretends to ignore.
Bias in AI is not limited to race in medicine. Studies show facial recognition systems misidentify women and people of color at far higher rates than white men, and hiring algorithms have systematically downgraded résumés from women. Forcing AI to ignore history and context means inequality will replicate, and the cost, particularly for patients already unseen by the medical system, will be measured in lives.
