Federal officials push AI into patient care despite safety risks
Government leaders and tech firms are integrating artificial intelligence directly into medical care to communicate with patients while experts warn that accurate advice can still harm behavior.
Accurate messages cause harm
Nominee Chris Klomp for deputy secretary of Health and Human Services regards AI in patient care as the agency's holy grail. Recent data indicates that over 300 million individuals seek health information from ChatGPT each week. Specialists caution that honest health information may alter actions to produce outcomes that harm well-being. Attempts by public health initiatives to stop binge drinking among college students and drug use by teenagers have not succeeded. Accurate public health campaign messages backfired even when they tried to stop these harmful behaviors.
Chatbots weaken vaccination intent
A study posed vaccination questions to ChatGPT, Gemini, and Claude by experts at the Vaccine Education Center. Each chatbot opened by describing chickenpox parties as popular or common before recommending vaccination. Research indicates that labeling an action as frequent may lead others to adopt it. One answer described vaccination as socially responsible but evidence suggests such altruistic appeals might even backfire. Two completely truthful answers produced opposite behavioral effects in the study conducted at the University of Pennsylvania.
Lawsuits focus on safety
Several states filed lawsuits against OpenAI in October, claiming ChatGPT offers hazardous medical guidance and deceives buyers regarding its safety. The lawsuits and company response focus on blatantly harmful content rather than accurate health advice that is unsafe. Anthropic faces a draft federal bill while the company says it will continue to improve the model. Officials in government maintain that artificial intelligence offers advantages since it excels at finding accurate medical data.
Reported by one outlet
Only one outlet has published this. Nothing here has been checked against a second report, so read it as that outlet's account and follow the link below for the original.
Reported by
1 independent outlet. Headline as published. Links open the original report.