Daily Briefing for October 30, 2026

Show Notes

AMA says doctors shouldn't be liable for AI errors they can't detect

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

Get the free email briefing

The American Medical Association on Thursday proposed matching liability for harm from clinical AI to who controls each part of it, saying physicians should answer for the judgment they can meaningfully exercise.

Developers would answer for keeping tools safe and effective, and hospitals and practices for how they select, govern and monitor them, according to the association's framework. Physicians should not be held liable for AI errors they cannot reasonably detect, Becker's reported from the accompanying issue brief.

In other news, a federal judge in Boston blocked a stricter federal test for which Medicaid enrollees count as medically frail under the new work requirements, and a Pew Research Center survey found 60% of U.S. adults would be uncomfortable with their doctor using AI for diagnosis and treatment.

In this episode

• AMA says physicians should answer only for AI judgments they can meaningfully exercise, not errors they cannot detect
• Judge blocks Medicare and Medicaid agency's stricter test for 'medically frail' patients weeks before Medicaid work requirements start
• 60% of Americans would be uncomfortable with their doctor using AI to diagnose and treat them, and only 3% would let AI diagnose alone
• Faster 0/1-hour troponin pathways do not shorten emergency stays or raise discharges vs. 0/3-hour, Lancet meta-analysis of 110,933 visits finds
• NYU Langone puts OpenEvidence inside its Epic record for 12.4 million outpatient visits a year, residents included
• Stamford Health pilots FDA-cleared AI that reads routine ECGs for hidden structural heart disease
• "It could be $10. It could be $10,000": hospitals struggle to forecast Epic's usage-based AI bills

What is Physician in the Loop?

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

Voice 1: This is Physician in the Loop, the daily briefing on AI in medicine. It's Friday, October 30th, 2026.

Voice 1: The American Medical Association says physicians shouldn't be held liable for AI errors they can't reasonably detect. That's the core of a liability framework its Center for Digital Health and AI released Thursday.

Voice 2: Give me an example of an error a doctor literally can't see coming.

Voice 1: The brief points to a triage model that deprioritizes a case the physician never sees at all. The framework ties liability to what it calls each party's, quote, authority, access to information and ability to prevent or mitigate harm.

Voice 2: So who actually answers for a bad outcome under this scheme?

Voice 1: Developers answer for keeping the tool safe and effective. Hospitals and practices answer for how they select, govern and monitor it. Physicians answer only for the clinical judgment they can meaningfully exercise, given the tool's known limits and what they knew at the point of care.

Voice 2: Does that let a vendor write contract language that shifts its own defects onto the doctor anyway?

Voice 1: The brief says no, vendor contracts shouldn't shift a developer's defects onto physicians. The AMA's president, Dr. Willie Underwood the Third, put it this way: "Patients must have meaningful avenues for recourse when clinical AI causes harm."

Voice 2: A framework is one thing. Whether a court actually uses it is another.

Voice 1: A federal judge handed a different kind of ruling to Medicaid patients this week.

Voice 2: A federal judge in Boston blocked a stricter test for who counts as medically frail under the new Medicaid work requirements, just weeks before those requirements start.

Voice 1: What was the stricter test actually requiring?

Voice 2: The 2025 law already exempts people who are blind or disabled, have substance use or disabling mental disorders, or have serious or complex medical conditions. CMS's rule had added a requirement that enrollees show their condition would, quote, significantly impair their ability to work. Judge Richard Stearns struck that down Thursday, ruling for 25 states and the District of Columbia.

Voice 1: Did he leave anything from the rule standing?

Voice 2: Two pieces. States must still reject frailty proof based on a diagnosis more than a year old, and can accept only one self-attestation per continuous enrollment period. Stearns wrote that enforcing the narrower impairment test could lead to disenrolling otherwise eligible vulnerable people, and he issued a permanent injunction.

Voice 1: So the underlying work requirement, 80 hours a month unless you're exempt, is still coming. It's just the definition of who's exempt that got easier.

Voice 2: A big study out of the emergency department landed the same day, and it's a quieter kind of news.

Voice 1: A Lancet meta-analysis of more than 110,000 emergency visits found that a faster, one hour troponin test doesn't actually get chest pain patients out the door any sooner than the older three hour version.

Voice 2: Faster test, same length of stay? What were the actual numbers?

Voice 1: Emergency stays ran about 314 minutes with the one hour pathway against about 308 minutes with the three hour one, basically the same. Direct discharge rates were close too, 67% versus about 66%. And 30 day death or new heart attack was low either way, under 1%, though the authors call that safety estimate imprecise.

Voice 2: So hospitals that switched to the faster test expecting shorter stays might not see it?

Voice 1: That's the implication. The researchers, led by Dr. Jasper Boeddinghaus, say hospitals weighing the switch should factor in their own local diagnostic and organizational setup rather than assume speed on paper translates to speed in practice.

Voice 2: And a reminder that the trials feeding this had a lot of disagreement between them, given how wide that heterogeneity got.

Voice 1: Patients themselves aren't sold on any of this, by the way.

Voice 2: A Pew survey found 60% of American adults would be uncomfortable with their own doctor using AI to diagnose and treat them.

Voice 1: Only 30% said they'd be comfortable. What about letting AI handle things on its own, no doctor checking?

Voice 2: For diagnosis, just 3% would accept AI acting alone, 47% wanted a human final review, and 41% wanted no AI at all. Reading scans got a bit more comfort, 6% alone, 60% with review. Insurance claim decisions were the least trusted: only 5% alone, and 46% wanted no AI involved at all.

Voice 1: Did people expect this to actually help or hurt them?

Voice 2: Split almost down the middle. 37% expected AI to make patients' health better, 36% expected worse. And 47% thought it would worsen the patient-doctor relationship.

Voice 1: What about confidence in the people actually deploying these tools?

Voice 2: Not great. Only 26% had a great deal or quite a bit of confidence that doctors would make sure AI serves patients' interests, and 42% had little or no confidence in hospital leadership on that same question.

Voice 1: That's a tough backdrop for a liability framework that leans so heavily on doctors exercising sound judgment. The patients in this survey don't sound convinced the judgment part is being exercised well at all.

Voice 2: Before we go, something to watch.

Voice 1: The AMA's Interim Meeting runs November 6th through 10th in Orlando. It's the first House of Delegates meeting since this liability framework came out.

Voice 2: So delegates could actually vote on something tied to what we just covered?

Voice 1: The post expects AI liability and payment resolutions to come up there. Nothing's passed yet, but this is where the framework either gains traction inside the organization or doesn't.

Voice 2: A framework released on a Thursday is one thing. Watching a room full of delegates argue over it in Orlando is where we'll find out if it has any teeth.

Voice 1: Sources and corrections are at physicianintheloop.org, where you can sign up for the free email briefing. This briefing was read by AI voices.