AI in Healthcare

Today's Stories

1. 1

   ECRI expands reporting to cover AI errors

   ECRI, a nonprofit patient safety organization, is expanding its reporting system to include errors involving artificial intelligence in healthcare. As AI tools are used more in clinical settings, mistakes involving them can harm patients; tracking these errors helps providers identify and prevent problems early.

   The expansion means healthcare organizations can now report AI-related incidents through ECRI's existing channels, potentially leading to more transparent AI oversight.

2. 2

   Nurses in Austin protest Palantir's role in healthcare

   Dozens of protesters gathered Thursday outside 8VC, a venture capital firm in Austin, to protest Palantir's AI technology in healthcare. The protest was part of a nationwide action organized by the National Nurses Union, held in cities including Palo Alto, Los Angeles, Chicago, and Washington, D.C. Monica Gonzalez, a National Nurses Union member, said Palantir's AI is used by HCA Healthcare through Timpani, a scheduling app, which sometimes short-staffs nurses. Protesters argue staffing decisions should be made by humans, not AI. Palantir responded that its software helps hospitals reduce sepsis deaths and places patients in beds more quickly.

   U.S. Rep. Greg Casar, D-Austin, joined the protest, urging the crowd to block an "AI oligarchy from Palantir." Jake Medwell, a co-founder of 8VC, said the firm has nothing to do with Palantir except that one of his partners co-founded it.

3. 3

   Nurses protest Palantir's AI in hospitals

   More than a hundred registered nurses protested outside Palantir's former Palo Alto headquarters on Aug. 27, 2026, and in seven other cities, calling for hospitals and officials to cut ties with the AI firm. National Nurses United says Palantir's AI tools are affecting staffing and care decisions, and that hospitals like HCA Healthcare and Mount Sinai Health System are using them despite concerns about the company's work with defense and immigration agencies.

   The union, representing over 225,000 members, is preparing for contract negotiations with HCA, Kaiser Permanente and other employers. Palantir earned nearly $2 billion in net income in the first half of this year.

4. 4

   Google DeepMind's AI Co-Scientist now runs lab experiments and writes papers

   Google DeepMind has expanded its multi-agent system Co-Scientist into a lab-integrated research partner. It now plans experiments, writes code, controls lab equipment, and generates scientific manuscripts, not just hypotheses. The system has delivered experimentally validated results across three disciplines—materials science, biology, and computer science. Reliability modules cut fabricated key results down to 4 percent, compared to 90 percent for a comparison system.

   In a fully autonomous computer science experiment, the system designed an AI architecture that outperformed six frontier models on health benchmarks after correction, but only showed a significant advantage over the baseline in one of nine categories under human evaluation. The gap between lab assistant and autonomous researcher remains wide.

5. 5

   AI-only care predicted to surpass doctors by 2030

   A JAMA article, led by Ezekiel Emanuel and Vinod Khosla, concludes that autonomous AI alone will likely surpass doctors and human-AI hybrids in five core medical tasks — taking histories, diagnosing, ordering tests, prescribing, and managing chronic disease — by 2030. The authors reviewed published AI-in-medicine studies since January 1, 2024. The paper argues that "humans in the loop degrade AI performance," so clinicians should step back. This challenges the prevailing hybrid model and has drawn criticism from AMA CEO John Whyte, who notes some studies were simulations and a February 2026 Nature article found patients struggled to use LLMs effectively. UCSF's Robert Wachter, though initially skeptical, concedes the argument is important, saying "there will be times when humans will muck up the performance."

   The paper predicts a "superior autonomous AI" will surpass humans using AI by 2030, a claim authors call "unsettling but seems probable." Neal Khosla expects AI will get regulatory permission to prescribe medicine in coming years. Meanwhile, concerns about "de-skilling" doctors are rising, as medical schools debate whether trainees should rely on AI tools.

6. 6

   Osaka University's AI detects oral cancer in 30 seconds

   Osaka University and colleagues received regulatory approval from Japan's health ministry for an AI diagnostic support system that detects suspected oral cancer in about 30 seconds. The AI model was developed with support from Nvidia, and Morita Tokyo Manufacturing handles manufacturing and sales. Oral cancer affects about 10,000 new patients annually in Japan, and early lesions can be mistaken for mouth sores, delaying detection. The AI analyzes oral images to identify precancerous lesions, mouth sores, and benign tumors, helping dentists decide whether to refer patients to specialists.

   The system, named Mucosight AI, is now approved as a 'program medical device.' Development began in 2017, with 40,000 oral images from seven hospitals used to train the AI. The university team aims to connect local dental clinics with advanced medical institutions for early detection.