Daily Update for September 30, 2026

Show Notes

Kennedy calls AI better informed than any doctor; McKinsey sees AI in up to 22% of outpatient claims; Oura postpones IPO; Penn Medicine adds OpenEvidence

The voices on this episode are generated, and so is the script, from the day's post at Physician in the Loop. Every fact in the post links to its source.

Read the post and its sources

Health and Human Services Secretary Robert F. Kennedy Jr. told the Make America Healthy Again Summit in Washington on Sept. 29 that artificial intelligence can give Americans a second opinion better informed than any doctor.

Kennedy also said OpenAI CEO Sam Altman had told him it would be "malpractice" to diagnose without consulting AI, Forbes reported. A day earlier, McKinsey & Co. estimated that AI could perform the work behind 16% to 22% of U.S. outpatient claims, from low-complexity office visits to imaging interpretation. Smart ring maker Oura postponed its IPO on the eve of trading, and Penn Medicine opened OpenEvidence to 10,000 clinicians.

In California, Gov. Gavin Newsom's deadline to act on four health-AI bills, including limits on AI marketed as therapy, falls today. A University of Florida study found that five AI evidence-search tools each missed large shares of the published evidence on a clinical question.

In this episode

• Kennedy tells MAHA Summit that AI gives a second opinion "much better informed than any doctor in the country"
• McKinsey estimates AI could perform 16% to 22% of U.S. outpatient claims, including 5% to 7% that are imaging interpretation
• Oura postpones its Nasdaq IPO a day before trading was to begin, citing IPO market "uncertainty"
• Penn Medicine makes OpenEvidence available to 10,000 clinicians across its seven hospitals
• No AI evidence-search tool tested found 12% of a gold-standard evidence set; single queries often returned none of it
• Five-variable model predicts intractable hypoxemia during one-lung ventilation in repeat lung surgery with validation AUROCs of 0.75 to 0.78
• Cooper pilots Microsoft's Dragon Copilot for nurses in New Jersey after licensing it for more than 1,300 clinicians
• Newsom's deadline arrives on four California health-AI bills, including SB 903's limits on AI marketed as therapy
• Senate health committee schedules no vote on FDA nominee Heidi Overton; STAT reports vote waits until after midterms

What is Physician in the Loop?

How technological innovation is changing medicine for patients and the people who practice on the front lines.

Nora: This is Physician in the Loop, the daily briefing on AI in medicine. I'm Nora Davis.

James: And I'm James Hudson. It's Wednesday, September 30th, 2026.

Nora: Robert F. Kennedy Jr. told the Make America Healthy Again Summit yesterday that AI can hand Americans, in his words, “a second opinion that is much better informed than any doctor in the country.”

James: He actually said better than any doctor. Not a good supplement to one, better than one.

Nora: That's the quote, from Forbes. Also today: McKinsey's outpatient claims estimate, Oura postponing its IPO, and Penn Medicine opening OpenEvidence to 10,000 clinicians. Kennedy also said AI can “free us from medical tyranny,” and that Sam Altman told him it would now be “malpractice” to diagnose or prescribe without consulting AI.

James: Altman said that to him, directly?

Nora: According to Forbes and The Hill, yes. Vice President JD Vance was there too, saying experts no longer hold a “monopoly on knowledge,” and that a company whose product harms people “should suffer consequences for it.”

James: That's at least an acknowledgment the technology can go wrong.

Nora: CMS Administrator Dr. Mehmet Oz said the health system isn't using AI enough, but he acknowledged its risks too. The Hill reported nine publicly traded companies, AI firms among them, sponsored the event.

James: Companies with a stake in the answer, in the room where the question got asked. Was there an actual policy announcement, or just the speeches?

Nora: No new policy. Just what got said from the stage.

James: Big claims, no paperwork behind them yet.

Nora: A day before that summit, McKinsey put out a report estimating AI could perform the work behind 16% to 22% of U.S. outpatient claims.

James: What kind of claims are we talking about?

Nora: Low-complexity office visits, 11% to 15% of claims. Diagnostic and imaging interpretation, another 5% to 7%, which McKinsey says works out to 700 million to 900 million claims a year.

James: That's the reading and dictating side, not anyone laying hands on a patient?

Nora: Right, McKinsey calls it “discursive care,” the reasoning and dialogue work: intake, triage, diagnosis, referral, follow-up. They say AI can handle a substantial share of that for 11 million of the country's 19 million health care workers.

James: Does that mean fewer jobs?

Nora: McKinsey says fee-for-service organizations risk losing low-acuity visit volume, but rising demand means the workforce overall isn't projected to shrink, according to Becker's.

James: So it's a claim about what AI could do, not what it's actually doing today.

Nora: That's their own caveat.

James: A follow-up on Oura's IPO. It postponed the whole thing, a day before trading was set to begin.

Nora: After all that oversubscribed demand we talked about?

James: Same company. Oura cited uncertainty in the IPO market and set no new date. It says it's profitable, with 5.7 million paid members, and expects revenue to grow 90% in the fiscal year that ends today. Those are the company's own figures.

Nora: Ninety percent growth and they're still walking away from the table.

James: CEO Tom Hale put it this way: “We have the luxury of choosing our moment.” Its registration with the Securities and Exchange Commission still hasn't been declared effective.

Nora: A confident way to describe backing out the day before the bell.

James: Penn Medicine is rolling OpenEvidence's clinical decision support tool out to 10,000 clinicians across its seven hospitals.

Nora: What's actually new about that, beyond the number?

James: They're also building customized tools with clinicians from the Botswana-UPenn Partnership, a 25-year academic collaboration. Dr. Farouk Dako of Penn Medicine said, “Effective clinical AI will be shaped by local expertise.”

Nora: And this ties into that Anthropic funding commitment we mentioned before?

James: Right, Anthropic is putting in credits, engineers and funding for a free version of the tool in about 100 countries. OpenEvidence itself says it now has more than 1.1 million verified U.S. clinicians, and it handled 42 million consultations in August. Company numbers, not independently verified.

Nora: Still, 10,000 clinicians at one system is a real number to watch.

James: Before we go, something to watch. Governor Newsom's deadline on four California health AI bills is today, and as of yesterday, LegiScan showed no action taken on any of them.

Nora: What do they cover?

James: AB 1979 would stop AI from performing licensed clinical work or directing unlicensed staff. AB 2575 would guarantee a right to override AI. SB 903 would bar advertising AI chatbots as therapy, and it passed the Assembly 71 to 4.

Nora: That's a wide margin. Anyone pushing for a signature?

James: Senator Steve Padilla's office says thousands of licensed mental health professionals signed a letter urging Newsom to sign SB 903.

Nora: Second thing: the FDA just closed a door device makers wanted opened.

James: The Harrison.ai one?

Nora: Yes. A final order denied Harrison.ai's petition to exempt experienced makers of radiology detection, diagnosis and triage AI from 510(k) review.

James: So those tools still have to go through that review, no exemption just for experience. Kennedy spent yesterday talking about AI outrunning doctors, and today the FDA kept that review requirement in place instead of loosening it.

Nora: Sources and corrections are at physicianintheloop.org. I'm Nora Davis.

James: I'm James Hudson. See you tomorrow.