An investigation led by RFK Jr. has been found to contain dubious citations and signs that point to the use of artificial intelligence in its composition. During a press briefing situated in the White House’s Roosevelt Room on May 12, 2025, Robert F. Kennedy Jr., who serves as the U.S. Health and Human Services Secretary, presented alongside President Donald Trump. The primary health project detailed in this conference, which lays out the administration’s healthcare plans for the American population, seems to exhibit AI-created ‘hallucinations.’ These include false references and invalid links, suggesting the administration’s endorsement of unscientific approaches towards public health.
The subject investigation, supervised by Health Secretary Robert F. Kennedy Jr., was shown to include entirely made-up reference sources. This is commensurate with AI systems’ trend to ‘ imagine’ details—a repeat and escalating issue adding fuel to the fire of false information spread on the internet and among the general populace. Specialists express concern over using AI for producing reports that carry dire implications for human health, given that AI does not uphold factual accuracy and holds susceptibility to the biases of those who program it. This document, however, forms part of the administration’s strategy pertaining to the health of countless American adults and children.
Steven Piantadosi, a professor of psychology and neuroscience at UC Berkeley, commented, ‘The issue with the AI we have today is that it isn’t reliable – its foundation lies in statistical correlations and dependencies, not truths.’ He continued, ‘It doesn’t understand the roots of truth and lacks a clear concept of a meticulously formulated logical or statistical case. It doesn’t grasp the notions of evidence or the weighting of different types of evidence.’
Georges C. Benjamin, the leader of the American Public Health Association, declared, ‘This report is not rooted in evidence. It should be discarded in all intents and purposes at this time.’ He continued his scathing review, ‘This cannot serve as the basis for any policy-making. It cannot even facilitate a serious conversation given the unreliability of its contents.’
The Trump administration attributed the false citations to ‘formatting glitches’, which doesn’t account for the invented citations and the telltale signs of AI involvement. Later, an edited version of the report with replacement for some offending citations was uploaded by the administration. Experts state that the report is saturated with misleading information, often making claims not substantiated by the references provided.
The report deliberately leaves out certain reasons for substandard health results in the U.S., such as the impacts of COVID-19, the absence of universal healthcare, or the reality that the leading death cause among children is gun violence. Additionally, it offers no viable solutions for these issues. Certain evidences are also selectively highlighted, such as the growing prevalence of mental health disorders in children, without noting that clinicians are getting better at identifying and diagnosing these disorders.
The report also demonstrates an element of hypocrisy in its recommendations. It spotlights the spread of ultra-processed foods and applauds the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) in one of its sections. The WIC program is commended for its ‘successful history in enhancing children’s health.’
Contrarily, the Trump administration, via the Republicans’ reconciliation bill, is planning extensive cutbacks to social initiatives which are predicted to negatively affect WIC’s enrollment rate. This move jeopardizes the benefits enjoyed by millions of beneficiaries, including pregnant women, postpartum mothers, newborns and children. These beneficiaries rely on WIC for access to fresh, healthy nutritional options.
Also evident in the report is a lack of concrete reference to realistic solutions to the existing health issues. It is not focused on developing structured and systematic strategies to tackle the complex health-related problems faced by the American public.
The reliance on an algorithm-based tool to compile serious health implications clearly showcases a disconnect in the approach towards public health. This move opens up a pressing conversation about the inclusion of AI in decision-making processes, especially those that affect the lives of large populations.
The fabricated citations and AI-based development of the health report bring into question the very integrity of the information presented. By integrating AI, a system whose credibility is already ambiguous, the administration puts hundreds and millions of Americans at risk.
The administration’s response to criticisms and subsequent publication of a revised document does little to quell the growing concerns. Claiming ‘formatting issues’ adds to the lack of transparency, further eroding public trust.
Ultimately, the administration’s apparent disregard for robust and evidence-based data in formulating this report manifests in its evident lack of scientific proofing. This could lead to significant negative implications on health-related policies and programs.
Furthermore, considering the potential adverse consequences of the administration’s substantial cuts to social programs, the push for artificial intelligence at the expense of credibility magnifies ongoing concerns. These concerns pertain not just to the future of American healthcare, but also to the ethical use of AI.
Overall, the entire episode reveals a pressing need for a more transparent and accountable approach to health report creation. Rigorously cross-verified and scientifically formulated reports are a necessity in shaping effective policies that truly benefit public health.