Noteworthy AI news: Ambient AI class action, AI-implicated patient harms, AI-conflicted psychologists, more

Every time an ambient AI vendor boasts about how many providers use its tool, a hungry lawyer gets a plum lead for a class-action lawsuit. 

And a lot of such lawyers are now on high alert for just such an opportunity to pounce. That’s because, in late November, a law firm representing as-yet stealthy plaintiffs filed suit against eight-hospital Sharp Healthcare in San Diego. Sharp’s mistake? Failing to “obtain all-party consent before recording confidential doctor-patient conversations as required under California’s strict wiretapping law (CIPA).” Legal eagles at Boston-based Fisher Phillips take an early look at the case in brief commentary posted Dec. 9.

• The plaintiffs are claiming that ambient AI documentation amounts to electronic eavesdropping—even if the vendor never “listens” in the human sense, Fisher Phillips partners Usama Kahf, JD, and Danielle Kays, JD,  report in the piece. Evidently, audio files were sent to and stored in the vendor’s cloud system, opening the door for the plaintiffs’ attorneys to argue that “simply capturing audio and sending it outside the organization, even for transcription, is enough for liability.” The plaintiffs also accuse Sharp of “not using encounter-specific verbal consent, pre-visit notices, on-screen or auditory indicators that recording was active, or written authorizations,” the attorney-analysts add.  
   
• The Fisher Phillips experts believe the Sharp suit will ripple well beyond healthcare. They recommend a handful of steps to be taken by any organization using ambient AI in any industry. These include things like “Rewrite vendor contracts now,” “Ensure vendors don’t take liberties” and “Build a fast, verifiable deletion workflow.”
   
• The attorney-authors also issue a disclaimer. “It’s important to note that these are simply allegations at this very early stage of the litigation, and we only have one side of the story,” they write. “Sharp hasn’t yet had an opportunity to respond. Regardless of whether these allegations are proven, however, the lawsuit provides a perfect opportunity for businesses in healthcare and beyond to review their practices.”
   
  • The brief is posted in full for free.

One good legal wrangle deserves another. 

Here’s the latest in a longish line of ruminations on who should—or will—be held liable when AI has something to do with a patient being harmed. It’s from the global firm Clyde & Co., which has been headquartered in London for nearly a century. The piece understandably doesn’t directly answer the central question it asks, but it succeeds at injecting a dose of common sense as only a proper Londoner can supply it.  

• “As well as the different legal frameworks potentially in play, one must also consider how AI is used—and whether this will influence where any liability lies,” the commentary points out. “Some have argued that, if the algorithm is influencing the decision or reaching the diagnosis, it could become a question of whether the clinician can understand or explain how the diagnosis was reached because, if they cannot explain this, then can they really be responsible if something has gone wrong?” Read the rest.

Psychologists are embracing AI for everything except client care. 

Could some have mixed motives for the avoidance? Some just might. A new survey by the American Psychological Association shows almost 40% of PhD therapists worried about losing some job duties to AI in the future. Meanwhile, only 8% use the technology to assist with clinical diagnosis and just 5% assign chatbots to patient-care duties. 

• The association’s announcement of the survey findings notes that overall AI use is up within the profession: Some 56% of psychologists reported using AI tools to assist with their work at least once in the past 12 months, up from 29% in 2024. But they’re also concerned about its growing role within their profession and beyond it. An unignorable 92% of respondents expressed concerns about the use of AI tools in psychology. Most common among these were potential data breaches, unanticipated social harms, biases in inputs and outputs, a lack of rigorous testing to mitigate risks and inaccurate output or hallucinations. 
   
• APA CEO Arthur Evans Jr., PhD, notes that psychologists are drawn to the field because they’re passionate about improving people’s lives. However, “they can lose hours each day on paperwork and managing the often byzantine requirements of insurance companies,” Evans adds. “Leveraging safe and ethical AI tools can increase psychologists’ efficiency, allowing them to reach more people and better serve them.”
   

Also worthwhile:

• AI is coming for healthcare, and it terrifies me. (Medium)
   
• Most people trust doctors more than AI but see its potential for cancer diagnosis (Society for Risk Analysis)
   
• AI is quietly transforming nursing (Psychiatrist.com)