Trove of Documents Exposes Secret Palantir Role in Medicare's AI Pilot Program
"I know how much these companies prefer to keep behind closed doors," said a former insurance industry executive. "Now we're learning that Palantir has been operating behind those doors, too."
The software firm Palantir has drawn scrutiny and recent public ire over its role in President Donald Trump's mass deportation campaign, the illegal US war in Iran, and Israel's genocidal assault on Gaza.
But a trove of documents obtained by the Electronic Frontier Foundation (EFF) reveals that the notorious tech company's tentacles have reached into the American healthcare system—including a widely criticized Medicare pilot program that uses AI models to review requests for medical services.
The documents include an application from Zyter, a healthcare technology company that the Centers for Medicare and Medicaid Services (CMS) chose to participate in the Medicare pilot program, formally known as the Wasteful and Inappropriate Service Reduction (WISeR) Model. Zyter's application mentions that Palantir "provides mission critical software infrastructure for CMS" and other government agencies, as well as "multiple" insurance companies. Zyter uses Palantir's software infrastructure, EFF noted.
Another document states that Palantir provides "AI-driven decision-making solutions" for "federal government entities and major payer customers such as Humana, Cigna, and Blue Shield of California."
The Center for Health and Democracy Education Fund, a nonprofit group founded by former health insurance executive turned whistleblower Wendell Potter, released a report Tuesday analyzing the documents unveiled by EFF, homing in on the "previously undisclosed claims of integration of Palantir Technologies’ products by major health insurers."
"As health insurers utilize AI more frequently, they are partnering with some of the nation’s largest technology corporations. Yet many of these relationships, including the technologies being used and the purposes they serve, remain largely unknown to the public," the report observed. "Given that these partnerships already exist, health insurers should immediately release to the public all vendor agreements involving the use of patient data."
The WISeR pilot debuted this year and has so far been rolled out in New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington state. News reports and internal documents show that the program was hastily implemented and has resulted in confusion and care delays for seniors. Vendors chosen to participate in the experiment earn a percentage of the estimated value of a denied claim, which critics say creates a perverse incentive to reject insurance claims.
The Center for Health and Democracy Education Fund demanded in its new report that insurance companies "immediately release to the public all vendor agreement involving the use of patient data." Without such information, the group warned, the public will not know key information such as the extent of AI involvement in coverage decisions.
"Patients should not have to wonder whether an undisclosed algorithm influenced their coverage determination, whether their medical information was shared with an outside technology company, or whether an automated system was evaluated for accuracy before affecting their care," the report concluded. "As AI becomes increasingly integrated into health insurance, it should not be allowed to make care-altering decisions, especially when it was designed by a company driven by profit rather than patient health."
Potter, a former Cigna executive who is now a vocal critic of the private insurance industry and supporter of Medicare for All, wrote in a series of social media posts on Tuesday that "I know how much these companies prefer to keep behind closed doors."
"Now we're learning that Palantir has been operating behind those doors, too," Potter wrote. "Is Palantir's technology helping insurers decide what medical care gets approved or denied? And why are we only learning about these relationships now?"
