Mental health clinicians at Kaiser Permanente staged a one-day strike in California on Wednesday, protesting Kaiser Permanente’s growing use of artificial intelligence tools and warning of risks to patient safety and their own jobs. The walkout involved up to 2,400 psychologists, social workers, and other therapists, with nurses and other unionized hospital staff joining picket lines across the Bay Area, Central Valley, and Sacramento, KQED reports. The National Union of Healthcare Workers said Kaiser has shifted initial mental health screening away from licensed clinicians to phone operators and an online questionnaire powered by AI, which they argue can miss patients at high risk of self-harm or in crisis.
Therapists say the company is also trying to scrap workload limits that protect time for existing patients and may steer more care to outside contractors. The union contends Kaiser has declined to put into the contract that AI will not replace therapists. Katy Roemer, a nurse in adult and family medicine, said the California Nurses Association shares the concerns and wants to ensure that humans provide care for other humans, per the AP. “Is AI going to benefit patients? Is AI going to benefit the people that work for Kaiser Permanente? Or is AI going to benefit the bottom line of the corporation?” she said. “So we want AI that’s transparent, that is allowing people to do their jobs.”
Kaiser disputes that it is using AI to make clinical decisions or to cut staff, saying it is expanding its mental health workforce and has nearly doubled it in the past decade in Northern California. The Oakland-based nonprofit says AI is being introduced to reduce administrative burdens and improve access to care, not to replace human judgment. “We see technology—and AI, in particular—as a way to support you in managing your practice,” the company told employees in a message this week. An expert on digital health at the University of Michigan said anxiety over AI is widespread in health care, per KQED, in part because decisions about deploying the technology are often made without front-line worker input.