A Harvard and Beth Israel Deaconess study published in Science on April 30, 2026 found OpenAI's o1 model beat physicians in six diagnostic experiments, correctly diagnosing 78% of textbook cases versus 30% for doctors.
A Nature Medicine study from Oxford, UCL, and the UK AI Security Institute ran 810 simulated conversations through nine AI chatbots and found mental-health risk builds gradually through 'Vulnerability-Amplifying Interaction Loops' rather than single bad replies.
Bayesian Health's TREWS platform, developed at Johns Hopkins, received the first-ever FDA clearance for a continuous AI sepsis monitor on May 12, 2026, after a study found it cut mortality 18% when clinicians acted on its alerts.
The FDA's April 2, 2026 warning letter to Purolea Cosmetics Lab is the agency's first enforcement action to explicitly cite inappropriate AI use, after the company let AI agents draft manufacturing records without human review.
Leica Biosystems announced FDA clearance on August 11, 2026 for Aperio iQC DX, the industry's first standalone AI quality control software for digital pathology, which detects up to 24% more slide-scan artifacts than manual review.
The FDA cleared Artera's ArteraAI Breast on May 6, 2026, the first digital pathology-based AI risk stratification tool for HR+/HER2-negative breast cancer, delivering same-day metastasis risk scores from existing biopsy slides.
The FDA cleared Neosoma's Brain Mets software on December 17, 2025, an AI tool that detects, measures, and contours an unlimited number of brain metastases from standard MRI scans for radiation therapy planning.
The FDA cleared Anumana's ECG-AI algorithm on April 8, 2026, the first tool authorized to flag cardiac amyloidosis from a standard 12-lead ECG, validated across 25,525 patients with 78.9% sensitivity and 91.2% specificity.
The FDA cleared UpDoc V1.0 on December 23, 2025, the first Software as a Medical Device built around a patient-facing large language model, for managing insulin dosing in adults with type 2 diabetes.
A federal judge has ordered UnitedHealth to turn over internal documents about its nH Predict AI tool, which plaintiffs say denies Medicare Advantage extended-care claims that are reversed on appeal roughly 90 percent of the time.