Google Cloud has launched Gemini for healthcare in U.S. preview, adding healthcare skills, read-only EHR connectors, and partner integrations to its Gemini agent for prior authorization, chart prep, payer review, and benefits questions. Read-only access protects the chart, but it also means Gemini prepares work that people and other systems still have to finish.
What Is Covered in This Article:
- Google Cloud launched Gemini for healthcare in U.S. preview, extending the Gemini agent it introduced at Gemini at Work with healthcare skills, connectors, a partner ecosystem, and a governed platform.
- Skills target provider work (longitudinal record synthesis, pre-visit chart prep, and prior authorization packets and appeals) and payer work (medical policy review and member benefits explanations).
- Google named 27 connectors, including read-only Epic and Oracle Health access, Optum eligibility tools in limited release, CMS coverage rules, TriZetto, Salesforce Health Cloud, and Microsoft 365, plus 16 systems integrators.
- Read-only EHR access lowers clinical risk but keeps Gemini an assembly and drafting layer, while Epic and Oracle Health own the chart and are building their own agents.
- Futurum data shows most AI builders run multiple model providers and are tightening agent governance, which makes model choice and audit controls central to the pitch.
The News: On November 10, 2026, Google Cloud CEO Thomas Kurian announced Gemini for healthcare in a post on the Google Cloud blog. The industry solution is built on the Gemini agent that Google introduced the prior month and has four components: skills for clinical and administrative tasks, connectors to EHR, claims, and operational systems, a partner ecosystem, and a governed platform with model choice and audit logging. Google says the platform is built to support HIPAA compliance and that customer data is not used to train its foundation models. HCA Healthcare, Northwell Health, Optum Insight, Blue Cross Blue Shield of Massachusetts, and the American Society of Clinical Oncology (ASCO) are working with Google Cloud on the capabilities, which are available now in preview in the U.S.
Gemini for Healthcare Reads the EHR
Analyst Take: Google Cloud is pitching Gemini for healthcare as the orchestration layer above the EHR, claims, and administrative systems that health organizations already run, without replacing any of them. It reads from them, assembles the evidence a clinician or reviewer needs, and drafts the output, whether a pre-visit chart summary, a prior authorization packet, an appeal, or a benefits explanation. Google says clinicians spend nearly 28 hours a week reconciling information, and the workflows it chose are where that time goes.
Futurum data backs Google’s choice of workflows. Futurum’s 2H 2026 AI Platforms Global Enterprise Decision Maker Survey (published October 2026) found that companies judge their own AI applications mainly on efficiency improvements (78.8%) and cost savings (59.5%), yet only 48.8% say AI ROI has met initial expectations, the lowest of eight survey periods. The same report names data integration as the main obstacle to moving agents into real work. Prior authorization and chart prep are integration problems with countable outcomes (hours saved, denials avoided, and days to decision), which gives Gemini for healthcare a clearer path to proving value than a general-purpose assistant has.
Read-Only EHR Access Limits What Gemini for Healthcare Can Finish
The most important design choice is that the Epic and Oracle Health connectors are read-only, with no write-back to the chart. That is a sensible choice for a preview. It protects the system of record, simplifies clinical safety review, and lets health systems pilot without changing EHR permissions.
It also caps how much work Gemini can finish. Google says the agent takes multi-step workflows through to completion, but for clinical work, completion still happens in the EHR, by a person, or through a separate submission path. Gemini assembles the packet; a human or another system files it. Epic and Oracle Health have both introduced AI agents that work inside the chart, and Microsoft offers Dragon Copilot for ambient documentation inside EHR workflows. The EHR vendors own the screen clinicians work in, and they can write back.
Google is betting that health systems want a cross-system layer that no single EHR vendor controls. I think that bet pays off for organizations running more than one EHR after mergers, but it is a harder sell to a health system standardized on Epic.
The Payer Side Is the Stronger Opportunity for Gemini for Healthcare
The payer use cases fit an outside orchestration layer better. Medical policy review and benefits explanations run on documents the health plan owns (policies, benefit booklets, and submitted clinical records), so read-only access matters less. The CMS Coverage connector grounds reasoning in National and Local Coverage Determinations rather than model memory, and the Medicare Advantage skill applies Medicare’s coverage criteria where a plan’s internal rule is stricter, in line with CMS limits on Medicare Advantage plans’ internal coverage criteria.
Regulatory deadlines add pressure. The CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F) set decision timeframes for impacted payers starting in 2026 and requires prior authorization APIs by January 2027. Payers must cut review time while documenting every decision, and Gemini’s traceable citations are aimed at that need. Blue Cross Blue Shield of Massachusetts describes Gemini for healthcare as the foundation for its reusable agent architecture, the most production-oriented statement among the customer quotes.
Google is also selling to both sides of prior authorization. Providers use Gemini to build stronger packets and appeals, and payers use it to review them. Optum, a connector provider and launch partner, sits on both sides as well. If both parties work from the same policy text and cited evidence, friction should fall. If each side’s AI simply produces more volume faster, the result could be more submissions, more appeals, and more automated denials. The shared-evidence design makes the first outcome more likely, but only production data will settle it.
Model Choice Is a Necessity for Google
Google frames model choice as a feature, and Futurum data suggests it is also a necessity. In the 2H 2026 AI Platforms survey, Google Gemini was in use and continuing at 39.3% of companies building AI applications (September 2026, n=511), behind Anthropic Claude (68.7%) and OpenAI GPT (62.4%), and 59.9% of model users run multiple providers in production. Google Cloud Vertex AI held a mean 12.5% of AI-specific cloud spend (April 2026, n=170), against 53.1% for three Microsoft products combined.
A healthcare solution limited to Gemini models would ask many buyers to give up models they already trust. Running partner models under one governance layer lets Google compete for the workflow even in accounts where it has not won the model. The same logic explains the ecosystem’s breadth: 27 named connectors, including Microsoft 365, and 16 systems integrators, including Accenture, Deloitte, KPMG, and PwC. Google is entering accounts where Microsoft and Epic are already embedded, so it has to meet buyers inside the tools they use.
Governance Will Decide Gemini for Healthcare Deals
Healthcare buyers will evaluate the governance model before the skills. Futurum’s survey found that formal monitoring of unauthorized access by AI agents rose from 42.8% to 52.2% in six months, the largest increase of any AI risk area tracked, and 75.2% of companies formally monitor data privacy or leakage (September 2026, n=600). Gemini for healthcare gives each agent its own identity, inherits each user’s role-based permissions, logs every interaction, and cites sources for every answer. Those are the right controls for an agent that reads protected health information across many systems.
Google also left out details buyers will need: pricing, accuracy or error rates for the skills, and any production results. The customer quotes use the language of exploration (“we’re exploring,” “we expect,” and “could”), which fits a preview but leaves the value case unproven.
My view is that Gemini for healthcare has the right architecture for a cautious industry, and the payer side is where it will show results first. Google should be judged within 12 months on whether HCA Healthcare, Northwell Health, and Blue Cross Blue Shield of Massachusetts publish production metrics such as denial rates, prep time per visit, and review turnaround, and on whether Epic and Oracle Health ever allow governed write-back. Without write-back, Gemini for healthcare risks becoming a well-governed research assistant that sits beside the systems where the work gets done.
What to Watch
- Pricing and packaging: Google has not said whether Gemini for healthcare is priced per seat, by consumption, or by outcome. Futurum’s September 2026 AI Assistants pulse found the share of companies capping metered AI consumption with hard limits or budgets set to rise from 21% to 38% within three months (n=100).
- EHR write-back: Whether Epic and Oracle Health open governed write-back to Gemini, or keep agentic action inside their own AI agents.
- Production metrics: Whether HCA Healthcare, Northwell Health, and Blue Cross Blue Shield of Massachusetts move from exploration to published results on denial rates, chart prep time, and review turnaround.
- Prior authorization APIs: How Gemini for healthcare works with the FHIR-based prior authorization APIs that CMS-0057-F requires of impacted payers by January 2027.
- Competitive response: How Microsoft, with Dragon Copilot and its share of AI cloud spend, and Optum, a partner that also sells its own AI tools, position against an orchestration layer that Google controls.
See the complete announcement blog post by Thomas Kurian on the Google Cloud website.
Sources
Declaration of generative AI and AI-assisted technologies in the writing process: This content has been generated with the support of artificial intelligence technologies. Due to the fast pace of content creation and the continuous evolution of data and information, The Futurum Group and its analysts strive to ensure the accuracy and factual integrity of the information presented. However, the opinions and interpretations expressed in this content reflect those of the individual author/analyst. The Futurum Group makes no guarantees regarding the completeness, accuracy, or reliability of any information contained herein. Readers are encouraged to verify facts independently and consult relevant sources for further clarification.
Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.
Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.
Read the full Futurum Group Disclosure.
Other Insights From Futurum:
Remediation Is Becoming a Software Delivery Capacity Problem
Google Agent Compute Engine Puts Cross-Cloud Governance to a Parity Test
Alphabet Q3 FY 2026: Cloud AI Growth Contrasts With Search Challenges
Author Information
Keith Kirkpatrick is VP & Research Director, Enterprise Software & Digital Workflows for The Futurum Group. Keith has over 25 years of experience in research, marketing, and consulting-based fields.
He has authored in-depth reports and market forecast studies covering artificial intelligence, biometrics, data analytics, robotics, high performance computing, and quantum computing, with a specific focus on the use of these technologies within large enterprise organizations and SMBs. He has also established strong working relationships with the international technology vendor community and is a frequent speaker at industry conferences and events.
In his career as a financial and technology journalist he has written for national and trade publications, including BusinessWeek, CNBC.com, Investment Dealers’ Digest, The Red Herring, The Communications of the ACM, and Mobile Computing & Communications, among others.
He is a member of the Association of Independent Information Professionals (AIIP).
Keith holds dual Bachelor of Arts degrees in Magazine Journalism and Sociology from Syracuse University.
