By Garrett Leahy & George Kelly

Clad in red shirts and holding signs reading, “Trust nurses, not AI,” hundreds of nurses from across California picketed outside Kaiser Permanente's San Francisco Medical Center on Monday to protest the use of artificial intelligence by the major hospital system—and the greater health care industry.

Nurses who rallied outside the hospital on Geary Boulevard as part of the protest organized by the California Nurses Association argued that the technology is untested and unregulated and said they fear it will make patient care worse.

“There are some uses of AI that are not acceptable,” Michelle Gutierrez Vo, a family medicine nurse, said during the demonstration.

The action coincided with the start of the KP International Integrated Care Experience conference in the Bay Area. The three-day event includes discussions on the use of advanced analytics and AI in health care, according to a public itinerary posted online.

According to the union, Kaiser Permanente is one of the earliest adopters of AI in the health care industry. The California Nurses Association, which represents 24,000 Kaiser nurses, is demanding that workers and unions be involved in the development of the technology and decisions on how AI is used in patient care.

“Nurses are all for tech that enhances our skills and the patient care experience,” Gutierrez Vo said in an earlier statement. “But what we are witnessing in our hospitals is the degradation and devaluation of our nursing practice through the use of these untested technologies.”

Cathy Kennedy, a neonatal intensive care nurse at Kaiser Roseville, said nurses understand that the new technology “is here to stay,” but she and her colleagues are concerned that Kaiser is using AI to cut costs and boost profits, rather than to improve patient care.

"We need to pause and really take time to say, is this truly what you want to do?" Kennedy said in a phone interview ahead of the protest. "We're questioning the validity of whether it's that important to continue increasing the amount of technology they are asking us to utilize as we're trying to take care of patients."

Kennedy, who has worked as a nurse for four decades, said some existing technologies already cause care delays, like electronic records that link staffing needs to real-time charting. If nurses don't log records immediately, the next shift could be short-staffed.

In an emailed statement, Kaiser said the company uses AI tools to empower nurses and that the technology does not have the power to make health care-related decisions.

“We have consistently invested in and embraced technology that enables nurses to work more effectively, resulting in improved patient outcomes and nurse satisfaction, and we will continue to do so,” the statement said. “As an organization dedicated to inclusiveness and health equity, we ensure the results from AI tools are correct and unbiased; AI does not replace human assessment.”

But Gutierrez Vo said Kaiser currently gives patients the option to talk to an AI chatbot in lieu of calling an advice nurse. The problem with that is the chatbot uses medical jargon as keywords to sort and prioritize patients, so if patients describe symptoms indicative of a heart attack but not with medical terms the AI is trained on, they might not be directed to seek emergency care in a situation when a real nurse would.

“The AI might direct them to the pharmacy, not to a doctor,” said Gutierrez Vo.

Kaiser nurses are arguing that any new technologies need to be properly evaluated before they are deployed. The union is planning to invoke bargaining rights to negotiate the hospital system’s use of any new AI tools before they are rolled out.

Nurses at other health care systems in California said AI tools at their hospitals are making their jobs harder, not easier.

Melissa Beebe, a nurse who takes care of cancer patients at UC Davis Medical Center in Sacramento, said the hospital currently uses AI to monitor patients’ vitals. The “bio button” tool alerts nurses if there are irregularities like an abnormal heart or breathing rate, but it often issues false alarms, Beebee said.

“The AI will say the respiratory rate is odd, but then I’ll look, and it’s fine,” Beebe said. “It’s not doing what it’s supposed to do, and we’re having to do double work.”

UC Davis Medical Center spokesperson Steve Telliano said Beebe’s concerns amounted to “fear-mongering,” saying that tool was introduced in May 2022 and allows patients’ vitals to be monitored while they are outside the hospital and to alert nurses of abnormal vital signs.

“The concern is that this [technology] will take jobs, it will not,” Telliano said. “It will make their jobs easier.”

But Beebe had another idea for how the hospital could make her job easier—and improve patient care.

“They could set up another oncology unit,” Beebe said. “But that wouldn’t save money.”