Speaking at the Association of Health Care Journalists Fall Summit in Los Angeles, Mishal Khan of the UC Berkeley Labor Center argued that AI in healthcare has benefits but also hidden pitfalls. She noted that digital technology has long been used to intensify work and degrade conditions, and that even without job losses, workers can still be harmed. Khan grouped AI uses into three categories: electronic monitoring, algorithmic management, and task automation, with potential harms including job de-skilling, lower wages, speed-up, discrimination, and suppression of organizing.
Deskilling was singled out as a major concern because AI taking over specialized tasks can narrow human expertise and threaten clinicians' scope of practice. Stanford's Tina Hernandez-Boussard offered a striking example from another session: medical residents sent to a Ugandan cardiac ward lost electricity and could not rely on AI for EKG readings, leaving them unable to synthesize the data confidently. Khan also flagged alienation from patients through automated handoffs, unresolved liability when AI recommendations lead to mistakes, and questions about what data ambient scribes collect on workers.
Khan predicted growing 'gigification' of healthcare jobs, with algorithmically managed nursing platforms that push workers to bid for shifts and could predict the lowest wage they would accept. Some states are already responding: Illinois and Vermont have passed bills prohibiting AI replacement of mental health workers, while California bars advertising AI chatbots as health professionals and is considering a measure to prevent workers from being forced to train their AI replacements. Khan's bottom line: 'AI can definitely have benefits for workers, but in the absence of real power, we're seeing more harms than benefits.'