Google Cloud launches Gemini for healthcare with HCA, Northwell and Optum as early users

Google Cloud on 10 November unveiled Gemini for healthcare, an agentic AI product for hospitals and health insurers that pulls together electronic health records, billing systems, clinical guidelines and Centers for Medicare & Medicaid Services datasets to carry out multi-step tasks². It is in preview, with launch customers including Blue Cross Blue Shield of Massachusetts, Hartford HealthCare, HCA Healthcare, Highmark Health, Northwell Health, Optum and Stanford School of Medicine².

What Google is offering

The product sits on Google's general Gemini agent and is packaged around prebuilt skills for jobs such as synthesising medical records, preparing for appointments, reviewing coverage policy and working out payer benefits, with customers able to write their own². Its record connections to Epic, through open.epic APIs, and to Oracle Health, through Ignite APIs for FHIR, are read-only: care teams can bring in active conditions, medications and recent labs, but the agent does not write back to the record². A NextGen Healthcare connection is in development for the coming months².

The connector list reaches well beyond the record. Google names lookups into CMS coverage data, the National Provider Identifier registry, RxNorm and ICD-10, eligibility checks through Optum, Waystar and Cognizant Trizetto, Suki's ambient AI scribe, GoodRx prescription pricing, and guideline content from the American Society of Clinical Oncology and BMJ Digital Health². Customers can route work across models from Google, Anthropic and OpenAI under one governance framework, with cheaper models handling high-volume administrative tasks². Google says it never trains its foundation models on customer data².

Thomas Kurian, chief executive of Google Cloud, said the problem is "the immense friction of making that data actionable across a highly regulated ecosystem"¹. At Stanford, deputy chief information officer Todd Ferris said researchers will be able to query an OMOP database in plain language instead of relying on SQL and an analyst².

What it means for the market

Google is the latest of the large technology companies to build a dedicated agentic AI product for health care, following Anthropic and OpenAI, which both launched specialised health tools this year¹. For health systems and payers, the pitch is consolidation: one governed agent layer over existing EHR, revenue cycle and office software, including Microsoft 365, ServiceNow and Salesforce's Health Cloud, rather than a separate tool for each workflow². That puts pressure on point-solution vendors in prior authorisation, eligibility and chart summarisation, although several of them, such as Infinitus, UiPath and Synthpop, have joined as partners².

Pricing will be offered either as per-user subscriptions or pay-as-you-go, and the product will be open to Gemini Enterprise Standard and Plus customers². A long list of systems integrators, including Accenture, Deloitte, KPMG and PwC, is building on it, which points to large enterprise deployments rather than departmental pilots².

Caveats

The launch rests on Google's own announcement and the early adopters' statements, several of which describe exploring the tool rather than results from using it². No figures on time saved, accuracy or cost have been published for the product itself, and the read-only EHR design limits what the agent can do inside the record². Google declined to say when Gemini for healthcare will move beyond preview to all enterprise customers¹.