Healthcare AI today: Cancer research rethought, Trump/Kennedy healthcare AI, ‘What’s best for the patient?’

News and views you ought to know about:

• AI has begun overhauling the means by which cancer research is conducted. And a new generation of AI-native researchers is leading the way. This next crop of scientists is AI-native, interdisciplinary and comfortable challenging assumptions. So reports Alicia Zhou, PhD, chief executive of the NIH-funded Cancer Research Institute, in a blog post. Encapsulating salient points made at a recent panel discussion, Zhou adds to the above observation: “By removing traditional barriers and giving emerging talent room to lead, institutions can accelerate innovation and move discoveries from bench to bedside faster than ever.” 
   
  • Other takeaways from the discussion as relayed by Zhou, who moderated the session: 1.) AI is poised to reshape every step of cancer drug development. 2.) The future hinges on the right kind of data—not just more of it. 3.) Immunotherapy is evolving, and AI can help unlock its full potential. 4.) Scientific creativity and collaboration are critical.
     
  • “We need better guesses,” panelist Alex Marson, MD, PhD, of the Gladstone Institutes and UC-San Francisco, told attendees. “How do we expand our imagination and hear hypotheses from AI that we wouldn’t have come up with ourselves?”
     
• The Trump Administration would like to train healthcare AI on patient data. The advance peek at its plan comes courtesy of Politico, which obtained a first draft of HHS’s Make America Healthy Again (MAHA) document. If finalized as is, the plan would let the National Institutes of Health under the leadership of Jay Bhattacharya, MD, PhD, build and maintain a digital platform for distributing and safeguarding the real-world data. True to the vision of HHS Secretary Robert F. Kennedy Jr., the draft report favors diet and exercise over medication for chronic conditions like diabetes and high blood pressure, Politico reports. 
   
  • The draft further calls for concerted FDA oversight of approved products over their lifecycles. Device categories would include products equipped with AI. In fact, the agency may decide to concentrate on those. According to Politico, the leaked document also says HHS would “research how AI could be deployed to assist with chronic disease care—including how it can be used to diagnose people earlier, personalize treatment and monitor patients as well as use predictive models to prevent hospitalizations.”
     
  • And while he’s at it, Kennedy will want to work with the FTC and the DOJ to crack down on social media influencers and consumer telehealth companies that mislead the public with slick promotions of prescription drugs. 
     
• The American Medical Association has compiled a list of 10 things every AI-adopting provider org should spell out in a policy paper. At the top of the list are definitions of common AI terms, descriptions of risks to patient safety and data security, and prohibited AI uses “such as entering patients’ personal health information into publicly available AI tools.” The list is just one part of what AMA calls its AI toolkit. More here. 
   
• Some Empire Staters like what their governor is doing to support innovation in healthcare with AI. It’s the Trump Administration they’re calling on the carpet. Which New Yorkers are we talking about? Those inclined to agree with Winthrop “Win” Thurlow. He’s executive director of LifeSciencesNY, a membership organization that serves as an advocate for New York’s medical device and life sciences industry. “Gov. Kathy Hochul’s pioneering Empire AI consortium is providing a blueprint for leveraging shared infrastructure and academic resources to promote AI research and development,” Thurlow states in an opinion piece published in the upstate-based Daily Messenger. “Unfortunately, decisions coming from the White House that chase researchers to other countries and a slate of bills introduced or passed in states like New York have the potential to dismantle this progress.” Hear him out. 
   
• Meanwhile some Golden Staters like what some of their elected officials are doing to strengthen an existing law for the AI age. The standing law prohibits unlicensed poseurs from presenting themselves as medical professionals. The bolstering measure is a bill that would explicitly extend such prohibition to anyone propagating AI-generated content and interactivity for that purpose. So far the bill has sponsorship from nine Democrats and no Republicans. Analyzing the new proposal, a lawyer with a New York City-based firm, Epstein Becker Green, asserts that “the days of operating in a regulatory gray zone are numbered.” The regulatory perimeter around healthcare AI, adds Alapp Shah, JD, MPH, “is tightening such that states may begin to legislate how AI can present itself to the public—not just what it does in the background.” Get Atty. Shah’s full take here. 
   
• What’s best for the patient? That’s the No. 1 question every user of clinical AI should be asking, case by case, as the technology continues to proliferate in healthcare. So maintains Björn Zoëga, MBBS, PhD, chief executive officer at King Faisal Specialist Hospital & Research Centre in Riyadh, Saudi Arabia. “The field is facing a patient-doctor-AI triangle dilemma—the patient needs to trust the doctor, who needs to trust AI,” Zoëga writes in a piece published Aug. 19 by Fast Company. “AI may be better and faster at identifying rare cancers or subtle diagnostic patterns, but only a doctor can consider a patient’s story, their social context, their fears—and translate a diagnosis into a treatment plan they can understand and embrace.”
   
• We have seen healthcare’s high-tech future, and it is in … wait. Robotic automation for administrative processes? Yes, that’s right. Robotic process automation, or RPA, “doesn’t diagnose cancer or save lives in the operating room,” the editors at Pymnts explain. But, till now, healthcare revenue management and payments processing “has been a labor-intensive, error-prone operation that can hemorrhage cash through administrative waste. In a sector where the business of care often collides with the practice of medicine, RPA innovations can increasingly help hospitals stay solvent, patients stay informed and clinicians focus on care rather than paperwork.” Get the rest. 
   
• From AIin.Healthcare’s sibling outlets:
   
  • Health Exec: Epic partners with Microsoft to roll out AI-powered clinical scribe
     
  • Health Imaging: AI as a second reader outperforms radiologists
     
  • Radiology Business: AI startup Better Medicine balloons fundraising to nearly $7M