Health AI in the 1st Week of February

ConcertAI (US): Debuted "Accelerated Clinical Trials" (ACT), an agentic AI platform designed to automate protocol design and patient matching. The goal is to cut study design time by 50% and reduce the mid-trial amendments that plague 76% of studies. [Link]

Oath Surgical (US): Partnered with NVIDIA to launch OathOS, a spatial AI platform for outpatient centers. It uses real-time video and audio analysis to reclaim the one month per year surgeons currently lose to inefficient technology. [Link]

Oracle (US): Unveiled the Life Sciences AI Data Platform, aggregating 129 million+ de-identified patient records. The unified environment is built to support agentic AI analysis across the entire drug development lifecycle, preventing data fragmentation. [Link]

NHS (UK): Launched a pilot at Guy’s and St Thomas’ combining AI risk assessment (Optellum) with robotic bronchoscopy. The pathway aims to shorten diagnosis times by replacing weeks of repeat scans with a single targeted biopsy for high-risk patients. [Link]

Ant Group (China): Scaled its "Ant Afu" AI agent to 30M monthly users within Alipay. By integrating Haodf’s 300k+ doctors and national insurance payments, Ant is transforming its "Super App" into a primary healthcare intermediary to bypass China’s strained hospital system. [Link]

Massive Bio (Global): Released "TrialRelay," a physician-facing operating system that uses AI to calculate the optimal "Green Route" for oncology referrals. It aims to eliminate the "black hole" of patient handoffs between intake and trial consent. [Link]

MHRA (UK): Issued guidance explicitly distinguishing "wellness" apps from regulated medical devices. The move attempts to curb the surge of unregulated mental health chatbots by defining strict "safe use" parameters. [Link]

Black Book Research (Global): A new 2026 benchmark report warns that hospital AI deployment is outpacing governance. Key risks include a widespread lack of "audit trails" and no routine monitoring for model drift or bias post-deployment. [Link]

UK Government (UK): Released a one-year update on its "AI Opportunities Action Plan," confirming it has met 38 of 50 commitments to lay the foundations for AI and secure homegrown innovation. [Link]

South Korea: Enforced the "AI Framework Act," designating AI for healthcare services and medical devices as "high-impact." This introduces strict risk management obligations, human oversight requirements, and penalties for non-compliance. [Link]

Imperial College London (The Lancet): The TRICORDER trial revealed a critical "human factor" flaw: while AI stethoscopes are accurate, the study failed to improve diagnosis rates because GPs frequently forgot or chose not to use the device. [Link]

Stanford Medicine (Report): The "State of Clinical AI 2026" audit serves as a reality check, finding that while frontier LLMs excel at reasoning, they often fail when confronted with the messy, incomplete data typical of real-world EHRs.[Link]

The Lancet (Study): Results from the MASAI trial (105k+ women) showed AI-supported screening reduced "interval cancers" by 12% and improved sensitivity (80.5% vs 73.8%), effectively reducing radiologist workload by triaging cases to fewer human readers. [Link]

Chung Shan Medical University (Taiwan): A study on LLM reliability revealed "authoritative bias," where models agreed with incorrect information if the prompt was phrased with authority. [Link]

Radiology Partners (US): Collaborating with Stanford Radiology’s AIDE Lab to validate AI tools across 3,400+ facilities. The partnership focuses on developing safety frameworks for monitoring AI in diverse, real-world clinical environments. [Link]

Innovaccer (US): Partnered with Snowflake to run its "Gravity" platform on the AI Data Cloud. Early adopters reported a 30% reduction in data integration timelines for enterprise AI workflows. [Link]

Group Health Cooperative (US): Became the first organization globally to deploy Epic’s new ambient AI charting tool in clinical practice, aiming to directly reduce the administrative burden on primary care physicians. [Link]

HealthEdge (US): Integrated Suki’s ambient AI into its GuidingCare platform. This extends "copilot" tech beyond doctors to payor-side care coordinators, streamlining member interactions. [Link]

Enlil (US): Partnered with Interlynk to integrate cybersecurity directly into the lifecycle of Software-as-a-Medical Device (SaMD). The tool generates "living" bills of materials (SBOMs) to ensure AI model transparency meets FDA standards. [Link]

Prenosis (US): Secured $40M (Series A + BARDA contract) to commercialize its FDA-authorized sepsis diagnostic and fund a major clinical trial for AI-guided treatment in acute care. [Link]

Automata (UK): Secured $45M (Series C) to scale its "operating layer" for life sciences. The funding will drive the adoption of robotics and unified data infrastructure to make physical labs "AI-ready." [Link]

Biorce (Spain): Raised $52.5M (Series A) to scale "Aika," an AI platform that generates regulator-ready clinical trial protocols in seconds. Backers include the CEOs of Revolut and Mistral AI. [Link]

Sword Health (US): Acquired Kaia Health for $285M to consolidate the MSK market. The deal provides immediate access to 70 million lives via Germany’s digital reimbursement pathway. [Link]

Spring Health (US): Agreed to acquire Alma to expand access to precision mental health care. The acquisition combines Spring Health’s employer-focused platform with Alma’s network of independent clinicians and national health plan relationships. [Link]