
U.S. Health and Human Services Secretary Robert F. Kennedy Jr. talks to guests in the East Room of the White House on May 22 following an event presenting a report from his Make America Healthy Again Commission. (Photo by Chip Somodevilla/Getty Images)
The 68-page report also includes a host of nonexistent or misleading citations of scientific reports and studies that point to the potential use of artificial intelligence in the drafting of the report, which HHS Secretary Robert F. Kennedy Jr. presented at the White House on May 22.
Nonprofit news publication NOTUS first reported on the false citations; other publications soon found additional errors. Updated versions of the assessment did not include correction notes and they introduced new errors, NOTUS found.
There are several clues pointing to the likely use of AI in the report, including formatting markers and similarities to past known uses of the technology. On a wider scale, AI has been popping up in unexpected places. Already, it's made its way into legal briefs, important documents and even declarations prepared by experts on artificial intelligence.
But experts who spoke with Law360 said they're less concerned about the potential use of AI in the report and more troubled about the seeming lack of checks that allowed the falsified citations to make their way into the published copy.
The fake citations also signal that the administration may be approaching public policy with predetermined conclusions instead of letting science lead it to the best policies, they added.
HHS spokesperson Andrew Nixon told Law360 in a statement that the "substance of the MAHA report remains the same — a historic and transformative assessment by the federal government to understand the chronic disease epidemic afflicting our nation's children."
"Under President Trump and Secretary Kennedy, our federal government is no longer ignoring this crisis, and it's time for the media to also focus on what matters," he said.
Nixon did not answer questions about the potential use of AI in creating the report or additional errors and the lack of correction notes in updated versions of the report.
Ivan Oransky, who teaches medical journalism at New York University's Carter Journalism Institute and is co-founder of Retraction Watch, which tracks retractions of scientific papers, said that while it's impossible to know definitively whether AI was used in the making of this report, it certainly resembles cases where AI has been used in the past.
Reports analyzing the fake citations in the MAHA report have found that some of the citations referenced studies authored by real experts in the field, but that did not exist in any scientific journal or were inaccurately summarized. Some references also included the term "oaicite" — a marker of the use of OpenAI in generating citations.
But the problem isn't necessarily that AI may have been used to help draft the report, Oransky said, but that the report was published with these incorrect citations in place.
"Either this was carefully checked over by people who didn't know what they were doing, or it wasn't checked over. Neither of those inspires any confidence," he said. "The problem is that the quality control on this document was obviously really subpar, and that's never good."
Oransky said that whether it was an intern or researcher or Kennedy himself, in the end, it doesn't matter much who actually drafted the report.
"When you see a strong signal that the work was just done sloppily, you have good reason to doubt what you read," he said.
The recently released MAHA report is designed to lay the groundwork for a second document, expected sometime in August, that will detail a strategy for "ending the childhood chronic disease crisis" in the country.
Experts previously told Law360 that based on the initial report, they could foresee the federal government making changes to vaccine policy, requiring stricter transparency standards for drug trials and creating policies targeting what the report described as the overprescription of medications in the U.S.
Richard Painter, a professor of corporate law at the University of Minnesota Law School and former chief White House ethics lawyer, told Law360 that false information has of course made its way into government reports in the past. But those were introduced by human beings and based on incorrect analysis or bias, he noted, not on citations to studies that didn't exist.
"I've never seen fake citations in a government report," he said.
But Painter added that he wasn't surprised when he heard about the fake citations.
"The Trump administration has been unique in actually attacking the notion of traditional concepts of scientific truth," he said. "There's not a determination to pin down the facts, get the facts right, before reaching conclusions."
Jolie Apicella, a partner in the litigation department and healthcare practice group at Wiggin and Dana LLP, said scientific papers are supposed to evaluate all the studies and information available about a particular topic and then build a report based on the findings, not have ready-made conclusions that researchers go "backing into" with studies they find that support that conclusion.
"Science doesn't work that way," she said.
The speed at which the report came together and the mistakes within it suggest that "they're finding conclusions first and in the process trying to find the support and the support might not be there," Apicella said.
She noted that the mistakes ultimately create doubts about the larger scope of the report, even if there may be credible findings.
The result now is that experts and critics are "scrutinizing it even more because it just becomes a less trustworthy document," she said.
--Additional reporting by Hailey Konnath and Steven Lerner. Editing by Alanna Weissman.
For a reprint of this article, please contact reprints@law360.com.