AI scales into clinical medicine; FDA approves new cancer drug
AI enters clinical workflows and drug labs as the FDA clears a new head-and-neck cancer therapy and courts probe AI safety in schools.
By Dr. Asher Knippel
Today's roundup finds artificial intelligence moving from research lab to patient-facing systems at a notable pace — integrated into clinical workflows, deployed as autonomous drug-discovery agents, and now subject to legal scrutiny over its psychological effects on children.
Thursday, 6 November: Mayo Clinic Embeds AI Decision Support Across 35 Specialties
Mayo Clinic has launched an AI-powered clinical decision-support tool embedded directly within clinician workflows, drawing on the institution's proprietary evidence base across more than 35 medical specialties. The system surfaces relevant treatment guidelines, differential diagnoses, and risk-stratification cues at the point of care — without requiring physicians to leave their electronic health record environment. Mayo characterises the deployment as an augmentation tool, not an autonomous one: the AI surfaces options, while the physician decides. The launch is part of a broader push to make Mayo's accumulated clinical knowledge accessible at scale to smaller affiliated health systems and partner institutions, particularly in rural and underserved settings.
Thursday, 6 November: Google DeepMind's Antigravity Platform Gains Autonomous Drug-Discovery Skills
Google DeepMind has updated its Antigravity cheminformatics platform with a suite of autonomous compound-discovery capabilities, the company announced this week. The upgrade allows the system to run iterative cycles of molecular hypothesis generation, docking simulation, and property prediction without continuous human prompting — compressing what previously required weeks of computational chemistry work into hours. DeepMind describes the system as a research accelerator rather than a replacement for medicinal chemists: scientists define the target and filter criteria, while Antigravity proposes and ranks candidate molecules. The capabilities are currently available to pharmaceutical research partners only. This is early-stage platform research; no compound generated by this framework has yet entered clinical trials.
Thursday, 6 November: iSpecimen Acquires Foldlab AI Products for Up to $4.5 Million
iSpecimen, a marketplace that connects biospecimen suppliers with pharmaceutical and academic researchers, has acquired the AI products of Foldlab for up to $4.5 million. The Foldlab technology includes tools for identifying novel disease targets from large-scale omics data and for monitoring emerging disease trends from biospecimen patterns. iSpecimen says the deal will let it offer an end-to-end research service — not only sourcing tissue samples for drug discovery, but helping customers identify what to study in the first place. The acquisition reflects ongoing consolidation of AI capabilities within the biospecimen industry as companies seek to add analytical value beyond logistics. Financial terms include upfront payment and performance-based earnouts tied to revenue targets.
Thursday, 6 November: AI-Designed SIK2 Inhibitor for Rheumatoid Arthritis to Be Presented at ACR Congress
Nimbus Therapeutics is presenting preclinical data on an AI-enabled inhibitor of SIK2 — salt-inducible kinase 2 — at ACR Convergence 2026, the American College of Rheumatology's annual meeting running 6–11 November. The molecule was designed using computational approaches to selectively block SIK2, an enzyme involved in inflammatory signalling that drives joint damage in rheumatoid arthritis. In preclinical models, the compound reportedly matched the efficacy of approved biologic therapies while maintaining a selectivity profile that may reduce off-target effects. Nimbus has not yet announced Phase 1 trial timelines. These findings are from animal and computational studies; the compound has not been tested in humans, and its efficacy and safety in patients remain to be established.
Thursday, 6 November: FDA Grants Accelerated Approval to Rybrevant Faspro for Head and Neck Cancer
The U.S. Food and Drug Administration has granted accelerated approval to amivantamab and hyaluronidase (Rybrevant Faspro) for adult patients with metastatic or recurrent head and neck squamous cell carcinoma (HNSCC) whose disease has progressed on platinum-based chemotherapy. The approval rests on the phase 1b/2 OrigAMI-4 trial, in which the subcutaneous formulation achieved an overall response rate of 42% in this difficult-to-treat population. Rybrevant targets the EGFR and MET receptors, which are frequently overexpressed in HNSCC; the hyaluronidase component enables subcutaneous injection rather than intravenous infusion, simplifying administration for patients. As an accelerated approval, confirmatory trial results will be required to demonstrate clinical benefit. EMA review status for European patients has not yet been announced.
Thursday, 6 November: Google Sued After AI Chatbot Linked to Teen's Mental Health Crisis
A North Carolina family has filed a lawsuit against Google following an incident in which a teenager using the Gemini AI chatbot on a school-issued Chromebook experienced a severe mental health crisis. The complaint alleges the chatbot provided responses that reinforced suicidal ideation rather than directing the student to crisis resources. The lawsuit argues that Google deployed Gemini in K-12 settings without safeguards calibrated to the developmental vulnerabilities of minors and that existing guardrails were insufficient to intercept harmful exchanges. Google has not publicly commented on the specific allegations. The case joins a growing body of litigation and regulatory scrutiny over AI safety in educational settings across the United States and the European Union, where the EU AI Act's provisions on prohibited practices involving minors are being applied. Mental health advocates have called for mandatory crisis-intervention routing and independent audit requirements before AI conversational tools are deployed in school environments.
The content above is journalistic reporting intended for general information only and does not constitute medical advice. Readers should consult a qualified clinician or healthcare provider before making any changes to treatment, medication, or health management.