Gemini for healthcare: Google Cloud AI agent for US hospitals, insurers

Google Cloud on Tuesday, 10 November 2026 launched Gemini for healthcare, a version of its Gemini AI agent built for hospitals, health systems and health insurers. It is available as a preview in the United States for customers on the Gemini Enterprise Standard and Gemini Enterprise Plus plans. Google says the tool is meant to take over paperwork-heavy tasks around care and coverage, such as pulling together a patient's records before an appointment, checking insurance policies and assembling evidence for prior-authorisation requests, while clinicians keep the final sign-off.

Health care runs on paperwork. Before a planned procedure in the United States, a hospital often has to ask the patient's insurer for permission, a step called prior authorisation. Staff gather notes, test results and codes from different computer systems, compare them with the insurer's rules and send a packet. If something is missing, the request is delayed or refused, and the patient waits. Insurers, on their side, read the same packets to decide what they will pay. Much of this is repetitive reading and matching, which is the kind of work that AI "agents" (software that can carry out a chain of steps, not just answer one question) are being built to do.

Google's pitch is that Gemini can sit across these silos. An electronic health record, or EHR, is the digital chart a hospital keeps for each patient; large systems such as HCA often run more than one. The company says Gemini can read across them, prepare a summary and point out gaps, so that people spend less time hunting for documents. The firm stresses that the AI drafts and coordinates, while doctors and staff still review and approve. That framing matters, because errors in medical records or coverage decisions have real consequences for patients.

The list of launch customers is notable because it includes both sides of the table: hospital groups such as Northwell, HCA and Hartford HealthCare, and payers such as Highmark, Optum and Blue Cross Blue Shield of Massachusetts. If both providers and insurers use similar tools to prepare and judge the same requests, the back-and-forth that slows approvals could shrink. Whether it does is not yet proven; the product is a preview, and the customers are at different stages, with HCA saying it is still exploring the tool.

For India, the launch is a signal more than a product. Gemini for healthcare is US-only for now, and India's insurance pre-authorisation process, including cashless claims under health policies and government schemes, works differently. But Indian hospital chains, insurers and the large number of health-technology and medical-billing firms based in India that serve US clients will watch closely, since agents that automate authorisation and records work could change the kind of back-office jobs those firms do.

Peepals timeline: Google introduced Gemini Enterprise as its workplace AI platform before this launch. On 10 November 2026 it added Gemini for healthcare as an industry version for providers and payers, in US preview, alongside a separate partner announcement the same day on using Gemini with NTT DATA to cut administrative friction in health care. Google has not given a date for general availability or for markets beyond the US.

Short-term: The seven named organisations and other eligible US customers can test the agent on records, policy and authorisation work.

Long-term: If the preview shows that AI can prepare and check authorisation packets reliably, hospitals and insurers could move more administrative work to AI agents, with people kept for review and decisions.

Who is affected: US hospitals and insurers, clinicians and administrative staff, patients waiting on approvals, and health-technology and outsourcing firms, including many in India, that do this work today.

Questions and Answers