The next time you’re scheduled for a medical examination, you might receive a call from someone like Ana: a friendly voice available to help you prepare for your appointment and address any pressing questions. Ana has a calm, warm demeanor typical of many nurses across the U.S., but unlike them, Ana is accessible 24-7 and can converse in multiple languages, ranging from Hindi to Haitian Creole.
This is because Ana is not a human but an artificial intelligence (AI) program developed by Hippocratic AI, among several new companies automating tasks that have typically been handled by nurses and medical assistants. AI is making its mark in healthcare, with numerous hospitals utilizing advanced computer programs to monitor patients’ vital signs, flag emergency situations, and initiate step-by-step care action plans—tasks traditionally managed by healthcare professionals.
Hospitals claim that AI assists nurses in working more efficiently, combating burnout, and addressing understaffing issues. However, nursing unions argue that this nascent technology is undermining nurses’ expertise and degrading patient care quality. “Hospitals have been anticipating a day where they can replace nurses with seemingly legitimate alternatives,” noted Michelle Mahon from National Nurses United. “The entire approach is designed to automate, de-skill, and ultimately phase out caregivers.”
Mahon’s organization, the largest nursing union in the U.S., has spearheaded over 20 demonstrations at hospitals nationwide, advocating for a say in AI use and protection from penalties if nurses choose to disregard automated recommendations. Concerns heightened in January when it was suggested that AI nurses, “as good as any doctor,” might provide care in rural locations. Moreover, Dr. Mehmet Oz, nominated to oversee Medicare and Medicaid, believes AI can free doctors and nurses from excessive paperwork.
Initially, Hippocratic AI advertised their AI assistants at $9 per hour compared to the $40 hourly rate for a registered nurse but has since shifted its messaging to emphasize carefully tested services. The company declined interview requests.
Hospital administrators have experimented with technologies to streamline costs and enhance care, integrating sensors, microphones, and motion-detecting cameras with electronic medical records to predict issues and direct nursing care preliminarily. Adam Hart, in the emergency room at Dignity Health in Nevada, experienced this firsthand when flagged by a sepsis alert—but found he was attending to a dialysis patient who needed careful fluid management, not blind adherence to protocol.
“You need to rely on your critical thinking—there’s a reason you’re a nurse,” Hart remarked. “Relying entirely on these devices is reckless.” Hart and colleagues acknowledge AI’s intention to streamline monitoring and responsiveness for nurses, but reality often translates to a flood of false alarms, flagging even basic bodily functions as emergencies.
Melissa Beebe, a cancer nurse at UC Davis Medical Center, expressed frustration at constant alerts: “It’s difficult to concentrate on your work amidst distractions that may not mean anything substantial.”
Michelle Collins, leading Loyola University’s College of Nursing, stressed that technology, regardless of sophistication, will overlook cues that humans like nurses naturally perceive. “Ignoring AI’s potential would be unwise. We should embrace its ability to augment our care but ensure it doesn’t supplant the human touch,” she said.
With an aging population and significant retirements, an estimated 190,000 nurse job openings are expected annually through 2032. This growing demand urges hospital management to adopt AI not to replace staff but to support them in gathering information and patient communication.
At the University of Arkansas Medical Sciences, AI assists in making hundreds of weekly calls to prepare patients for surgery during limited hours without incurring overtime costs. Since January, AI provided by Qventus has facilitated communication between patients and healthcare providers while managing medical records for human staff.
While Qventus contributes administratively, other developers envision a deeper role for AIIsraeli startup Xoltar collaborates with the Mayo Clinic on an AI assistant that engages patients in cognitive management of chronic pain and develops avatars aiding in smoking cessation, detecting non-verbal cues like facial expressions.
Health experts caution that although AI benefits patients who are proactive about their care, it may not yet address the needs of very sick individuals consuming most healthcare resources. “Whether chatbots fit these more critical circumstances warrants careful consideration,” noted Roschelle Fritz from the University of California Davis School of Nursing.
