RFK Jr.'s AI Army: When Health Meets Hype

Robert F. Kennedy Jr. has discovered artificial intelligence. During a recent stop in Nashville, the Health and Human Services secretary declared that his agency is "leading the federal government in driving AI into all of our activities." An army of bots, he promised, will transform medicine, eliminate fraud, and put a virtual doctor in everyone's pocket.

The reality is more complex—and revealing.

The AI Inventory: Revolution or Routine?

Last month, HHS published an inventory of roughly 400 ways it's using AI. The list reads less like a medical breakthrough and more like a corporate efficiency manual. The department is deploying chatbots to generate social media posts, redact public records requests, and write "justifications for personnel actions." One entry simply reads "AI in Slack"—referring to the workplace communication platform.

The agency's new RealFood.gov website, which promotes Kennedy's dietary vision, features a chatbot promising "real answers about real food." Click it, and you're redirected to xAI's Grok chatbot in a new window—hardly the seamless integration one might expect from an "AI revolution."

Many applications are undeniably mundane: managing electronic health records, reviewing grants, summarizing scientific literature. There are multiple IT-support bots and AI search tools scattered throughout the 400-item inventory.

The Staffing Connection

The proliferation of back-office AI tools coincides with a less discussed reality: HHS has fired or accepted voluntary buyouts from thousands of staff over the past year. The agency's database explicitly cites "staffing shortage" as the reason why the Office of Civil Rights is piloting ChatGPT to identify patterns in court rulings involving Medicaid.

This raises an uncomfortable question: Is Kennedy's AI push genuinely about innovation, or is it administrative automation designed to compensate for a hollowed-out workforce?

Sponsored

Advertise with Us

[email protected]

Sponsored

When AI Goes Wrong

The technology's limitations are already showing. FDA employees describe Elsa, the agency's custom AI tool launched last June, as "quite bad" and prone to failure. One staffer asked Elsa to look up a three-digit product code in the FDA's public database—it returned the wrong answer. Another tried using it to evaluate a food safety report; the AI "processed for a moment and then said 'yeah, all good,' when I knew it wasn't."

The problems extend beyond individual tools. HHS's landmark "Make Our Children Healthy Again Report" was initially riddled with fake citations that appeared to be AI hallucinations. The White House attributed this to "formatting errors," but the incident highlights the risks of deploying unproven technology across critical health infrastructure.

The Bigger Picture: Healthcare's AI Moment

Kennedy's embrace of AI reflects broader trends in healthcare. Doctors spend more than a third of their time on administrative tasks—writing notes, reviewing charts, processing insurance claims. If AI can automate even a fraction of this work, it could free up healthcare workers to spend more time with patients, addressing the chronic shortage of medical professionals.

Major tech companies are betting heavily on this vision. OpenAI and Anthropic recently launched healthcare products, while hospital networks nationwide are piloting AI tools for everything from diagnostic imaging to drug discovery. Google DeepMind'sAlphaFold, which won a Nobel Prize for protein folding research, is now used by researchers worldwide, including those at HHS.

The Political Dimension

But Kennedy's AI deployment isn't purely about efficiency or medical advancement. The inventory reveals at least one openly political application: HHS is using AI to identify positions violating President Trump's executive orders on "Ending Radical and Wasteful Government DEI Programs" and "Defending Women From Gender Ideology Extremism."

This signals how AI tools, ostensibly neutral, can become instruments of political enforcement. When Kennedy talks about using AI to "eliminate fraud," the question becomes: fraud according to whom, and by what standards?

The Promise and the Peril

Some HHS applications do show genuine promise. A tool used by federal and local health departments analyzes grocery store receipts from people with suspected foodborne illnesses, searching for commonalities that could prevent outbreaks. Other projects involve studying malaria genomes and accelerating drug safety reviews—work that could genuinely advance public health.

Yet the gap between Kennedy's revolutionary rhetoric and the mundane reality of most applications is striking. Many tools seem designed more to paper over staffing cuts than to unlock medical breakthroughs.