AI is exacerbating Americans’ distrust of health care

After OpenAI and Anthropic launched dedicated healthcare initiatives in January, a study published in February found that OpenAI’s ChatGPT Health had a 50% error rate, incorrectly recommending that care be delayed in emergency testing cases half the time.

This error rate, which had not been identified before the app’s deployment, is a symptom of a larger problem: the rapid adoption of AI systems by health systems and insurers, often skipping critical tests to determine how well these systems work and how safe they are for patients. This drive to develop AI in healthcare intensifies an existing crisis of trust.

The decline in trust in health care in the United States continues and has been worsened by institutional responses to the Covid-19 pandemic. A national survey of more than 443,000 U.S. adults found that trust in doctors and hospitals fell more than 30 percentage points between 2020 and 2024, from 72% to 40%, with declines across several sociodemographic groups. For Black, Latinx, and Indigenous communities, this collapse adds to a pre-existing medical mistrust rooted in a legacy and ongoing history of medical racism in the U.S. healthcare system. Research shows that patients who distrust their health care providers are more likely to delay care, including preventative screenings, and discontinue medications, and that these trends are associated with higher rates of hospitalization and premature death.

The documented harms of AI compound this mistrust. For example, a widely cited algorithm affecting approximately 200 million Americans systematically underestimated the health status of black patients, after using medical costs as a measure of illness. Patients were unaware that this tool was used to determine the level of their care. Medicare Advantage insurers used AI tools that helped double their denial rates for older patients; approximately 75% of denials were overturned on appeal, but less than 1% of patients appealed. The federal government has since launched a pilot for AI-based prior authorization in traditional health insurance in six states.

Healthcare, which is worth $5.3 trillion, or 18% of GDP in 2024, is heavily invested in by the AI ​​industry. U.S. healthcare organizations spent $1.4 billion on AI tools in 2025, nearly three times what they spent the previous year, for a range of functions including medical image analysis and billing and documentation automation. In addition to potential profits, the sector also provides what AI companies need to operate and, in many cases, to build and improve their systems: data, and a lot of it. This includes data in the form of electronic health records, insurance claims, diagnostic images, and genetic profiles of hundreds of millions of Americans, often collected without meaningful transparency about how it will be used and without input from patients and communities.

Data shows that the rapid adoption of AI in healthcare is compounding the distrust Americans already have in our healthcare system. A February 2025 study of more than 2,000 Americans found that 66% said they had little confidence in their healthcare system to use AI responsibly, and 58% said their healthcare system would ensure that an AI tool would not harm them.

Neither AI knowledge nor health knowledge changed these results. The most important predictor was how much trust a person already had in the healthcare system.

In a nationally representative survey, most patients said they wanted to know when AI was used in their diagnosis and treatment. Yet no federal law requires disclosure, and only a handful of states currently have laws to address this problem. When patients are not informed about what is happening to them or their data, and no one is required to share this information with them, it affects all patients, but particularly those communities with the least trust to lose.

Patients who have experienced discrimination in healthcare are significantly less likely to trust healthcare systems to use AI responsibly. Deploying AI systems without meaningfully involving patients and communities in decision-making only repeats the pattern that led to distrust in the first place.

What needs to change is who contributes to decisions about how AI tools are purchased, governed and used. Patients and community members need formal decision-making roles, not just advisory positions. Health systems and insurers must publicly report their performance, including across different racial/ethnic groups, before AI tools are deployed. Patients should be informed clearly and in advance when AI is used in their care. These are the basic conditions for a reliable system.

Health systems and businesses can make different choices, choices that earn the trust of their patients and the communities they serve. They have the ability to act quickly. The hardest work is done at the pace of trust. This means that patients and community members have a say before these systems are even purchased, not after harm has been caused.

Oni Blackstock, MD, MHS, is a physician-scientist, founder and executive director of Health Justice, and a Public Voices researcher on technology in the public interest with the OpEd Project.