AI News Monitor

A sweeping health report released last week by Health Secretary Robert F. Kennedy Jr., which sharply criticized America’s food supply, pesticides, and prescription drugs, included citations to studies that do not exist and appears to have used artificial intelligence to generate references, according to experts and a review by The Washington Post.

The report, titled Make America Healthy Again (MAHA), cited hundreds of studies to support its claims. But an investigation by NOTUS found that some of those studies could not be located, and experts later identified telltale signs that portions of the citations were generated by OpenAI’s chatbot. Some references included “oaicite” markers in their URLs, a strong indicator of AI involvement.

“This is not an evidence-based report, and for all practical purposes, it should be junked at this point,” said George C. Benjamin, executive director of the American Public Health Association. “It cannot be used for any policymaking. It cannot even be used for any serious discussion, because you can’t believe what’s in it.”

The Post found that at least 21 links in the original version of the report led to dead ends. By Thursday, the document had been quietly revised to remove the “oaicite” references, though experts said the errors undermined its credibility.

The Department of Health and Human Services dismissed the issues as “minor citation and formatting errors” that had since been corrected. Andrew Nixon, a department spokesperson, defended the report as “a historic and transformative assessment” of chronic disease in American children.

White House press secretary Karoline Leavitt also stood by the report, calling it “backed on good science” despite acknowledging “formatting issues.”

But critics, including former FDA official Peter Lurie, called the reliance on AI-generated citations “shockingly hypocritical.”

“They’re wrapping themselves in the shroud of scientific excellence while producing something sloppy and shoddy,” Lurie said.

The controversy raises fresh concerns about the administration’s use of scientific rigor in policymaking—and whether AI’s role in research could further erode public trust in government reports.