Latest updates · 21 October 2026
In the full issue →US regulators are shaping how generative AI medical devices will be judged: an FDA digital-health chief says they may need to be assessed more like physicians, and the FDA has posted worked examples of AI devices that update themselves under pre-approved plans.
Research, industry & policy
- FDA official: GenAI tools may need physician-style assessment, MedCity News · 20 Oct 2026: At AdvaMed's MedTech Conference in Boston, Rick Abramson of FDA's Digital Health Center of Excellence said the agency received more than 300 comments on its GenAI device discussion paper and intends to draft policy.
- FDA outlines AI device modification scenarios under PCCPs, FDAnews · 20 Oct 2026: Two hypothetical examples show a hospital-specific sepsis alert being fine-tuned on local data and a cardiac arrhythmia detector being retrained automatically, both within predetermined change control plans that avoid a new submission per change.
Journals & preprints
- When to Stop Asking: Factor Coverage Bias and LLM Determinability Judgments with Underspecified Information in Medicine, arXiv 2610.25819 · 19 Oct 2026: Clinical decision-making often begins with underspecified information and requires not only gathering missing information but also knowing when to stop.
- CODA: Cohort- and Drift-aware Foundation Model for Multimodal Clinical Reasoning, arXiv 2610.25871 · 19 Oct 2026: Recent advances in multimodal foundation models have demonstrated strong performance across diverse clinical tasks.
Conferences, workshops & events
- CAHOTECH 2026, 14–15 Nov 2026 · Visakhapatnam (AMTZ Campus): Healthcare technology conclave: digital health, AI, automation and robotics.
- HLTH USA 2026, 15–18 Nov 2026 · Las Vegas (Venetian Expo): Health-innovation event themed 'The Age of Intelligent Healthcare'.
- Global Digital Health Summit 2026 – AI in Action, 20–21 Nov 2026 · New Delhi (Bharat Mandapam): Theme: Implementing Artificial Intelligence – AI in Action.
Why it matters
How regulators test open-ended AI will decide which clinical tools reach patients and how quickly they can be updated safely. News summary only, not medical advice.
