TL;DR
Get health and wellness essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
At the Association of Health Care Journalists Fall Summit in Los Angeles, UC Berkeley researcher Mishal Khan warned that AI in healthcare can harm workers even when it does not eliminate jobs. She highlighted risks including loss of clinical skills, monitoring, reduced autonomy and algorithmically managed work; researchers and lawmakers are examining how to respond.
AI can harm healthcare workers even when it does not take their jobs, UC Berkeley Labor Center researcher Mishal Khan said at the Association of Health Care Journalists Fall Summit in Los Angeles on October 13. She warned that workplace systems may erode clinical skills, increase monitoring and reduce workers’ control over their jobs, raising questions about how automation affects both staff and patient care.
Khan, a senior researcher in the Technology and Work Program at UC Berkeley’s Labor Center, described three broad workplace uses of AI: electronic monitoring, algorithmic management and task automation. Monitoring can include cameras used to track workers or check compliance with safety procedures. Algorithmic management can analyze calls between employees and customers and prompt workers to change how they respond. Task automation shifts particular parts of a job to software.
In Khan’s account, potential harms include job loss, deskilling, lower wages, work speed-up, discrimination and reduced autonomy. Her study of AI’s potential effects in healthcare identified deskilling—workers losing specialized abilities as software takes over tasks—as a particular concern, including the possibility that it could affect healthcare workers’ scope of practice.
Stanford biomedical informatics professor Tina Hernandez-Boussard described a case involving medical residents sent to support a hospital cardiac ward in Uganda. When the hospital lost electricity and AI tools were unavailable to read electrocardiograms, the residents were not confident synthesizing the information to make recommendations, she said. The example was presented as a concern about dependence on automated support, not evidence that AI alone caused the residents’ difficulty.
Clinical Skills and Worker Control
The concerns reach beyond whether technology reduces staffing. If clinicians have fewer opportunities to practice tasks that software handles, their ability to work when systems fail could be affected. The Uganda example illustrates that concern, although the account does not establish how common such dependence is or whether residents’ training and other circumstances contributed.
Workplace systems may also shape how care is delivered. Khan said automated patient handoffs raise concerns among workers about patient relationships, while ambient documentation tools prompt questions about what information they collect from staff and how it might be used. If data gathered for one purpose later inform worker evaluations or management decisions, the technology could affect privacy and professional autonomy. The report does not say that this use is already occurring across healthcare.
These issues matter to patients as well as employees. Clinical skills, clear responsibility for decisions and continuity in patient relationships can affect care. Khan raised a liability question if an error follows an AI recommendation, but the summit report does not specify a legal answer. The effects are likely to depend on the technology, workplace rules and oversight in place.
As an affiliate, we earn on qualifying purchases.
Labor Concerns and State Measures
Khan placed healthcare AI within a longer history of digital tools used to monitor workers and intensify work. The practices she described are not limited to hospitals: monitoring drivers and coaching customer-service employees are examples of how employers can use technology to track behavior and guide work. Her argument was that similar arrangements may expand as AI is adopted in healthcare.
She also warned of the possible “gigification” of healthcare jobs. In her example, a nursing platform could use algorithms to allocate shifts through bidding and predict the lowest wage a nurse might accept. This was a potential scenario she described, not a reported finding that such a system is already widespread.
Khan cited policy activity in several states. She said Illinois and Vermont had passed bills prohibiting the replacement of mental health workers with AI, and that California law bars companies from advertising an AI chatbot as a health professional. She also referred to a California bill under consideration that would prevent workers from being required to train their replacements, including AI systems. The report did not provide bill numbers, implementation details or the status of that proposal beyond saying it was being considered.
“Even if they don’t lose a single job to AI, it doesn’t mean there are no harms to workers.”
— Mishal Khan, senior researcher at the UC Berkeley Labor Center
clinical skills training simulators
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Questions About Use and Accountability
The summit report does not quantify how many healthcare workers currently use the systems Khan described, how often they cause harm or how AI adoption has affected wages, injury rates or clinical performance. Her warnings identify possible risks and concerns; they do not establish that every AI tool produces those effects.
It is also unclear what data ambient scribes collect about workers, whether employers use those data to make employment decisions, and how responsibility would be assigned when an AI recommendation contributes to an error. The report gives no details about safeguards or accountability rules adopted by particular health systems.
The status and scope of the California proposal Khan mentioned, as well as the practical effects of the measures passed in Illinois and Vermont, were not detailed. The account also does not establish whether the Uganda incident is representative of medical training elsewhere.
AI healthcare workflow management tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Policy and Workplace Decisions Ahead
The next developments will depend on how health systems adopt AI and whether lawmakers set rules for its use. The California bill described by Khan was still under consideration at the time of the report; its progress and final provisions were not provided. The measures she cited in Illinois and Vermont, and California’s restriction on advertising AI chatbots as health professionals, are areas to watch as implementation and enforcement unfold.
For healthcare employers, the issues raised point to decisions about worker monitoring, the use of data from documentation systems, training for tasks supported by AI and responsibility when automated recommendations are involved. The summit report does not announce a new policy or industry-wide standard. Khan’s warning is that these workplace choices can affect workers even if no jobs are eliminated.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did Mishal Khan warn about healthcare AI?
Khan said AI can harm healthcare workers without eliminating jobs, including through deskilling, monitoring, work intensification and reduced autonomy. She framed these as potential risks of how employers use the technology.
What is deskilling in this report?
Deskilling refers to workers losing specialized abilities when software or other automation takes over tasks they would otherwise perform. Khan said it was a particular concern for healthcare and could affect workers’ scope of practice.
Did the report say AI caused the residents’ difficulty in Uganda?
No. Hernandez-Boussard said residents were not confident synthesizing information after a power outage made AI unavailable to read electrocardiograms. The report presents this as an example of concern about dependence on automated support, not proof that AI caused the difficulty.
What state measures did Khan cite?
Khan said Illinois and Vermont had passed bills prohibiting replacement of mental health workers with AI, and California law bars advertising an AI chatbot as a health professional. She also said a California bill about requiring workers to train their replacements was under consideration.
Is AI already replacing healthcare workers on a large scale?
The report does not provide data on the scale of job replacement or AI adoption in healthcare. Khan’s argument is that labor harms can occur even when workers do not lose their jobs.
Source: rss
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
