Google Cloud Launches Gemini for Healthcare to Automate Hospital Operations

On November 10, 2026, Google Cloud introduced Gemini for Healthcare, a specialized AI platform for medical organizations and insurance companies. The product allows for the integration of information from electronic medical records, insurance systems, and clinical directories to automate document preparation and query processing. The developers aim to reduce the time healthcare workers spend searching for information and handling administrative tasks.

Gemini for Healthcare is built on Google's corporate AI agent and works with existing information systems through special integrations. Participants in the launch include Northwell Health, Highmark Health, HCA Healthcare, Optum, Blue Cross Blue Shield of Massachusetts, Hartford HealthCare, and Stanford School of Medicine. The solution is currently available in preview mode, and its actual effectiveness will be evaluated during implementation.

Capabilities of Gemini for Healthcare

The platform combines several tools for the daily work of doctors and insurance specialists. It can gather medical history from various sources, prepare a summary before a patient visit, identify discrepancies in medical documents, and assist with pre-authorization of treatment with insurers. Other features allow for checking insurance coverage conditions and calculating estimated patient expenses.

For working with medical records, Google has provided integrations with Epic and Oracle Health. Through these, Gemini receives information on diagnoses, prescribed medications, test results, and specialist visits. The system can also access data from the U.S. Medicare program and medical guidelines, including materials from the American Society of Clinical Oncology (ASCO) and BMJ.

In a demonstration for Fortune, Google representatives showed how Gemini compares a patient's electronic record with a report from an external doctor. In 35 seconds, the system gathered key health information, identified missing test results, and additional allergy information not present in the record. Each identified discrepancy was accompanied by a source link for medical professionals to verify the findings.

The AI does not independently alter the original records in Epic and Oracle Health. Access is organized in read-only mode, and updating the medical record remains the responsibility of authorized personnel. Google also claims to provide user rights management and tools for auditing system actions.

Why Google is focusing on administrative processes

One of the main issues in American healthcare remains the costs associated with medical insurance and document processing. According to a KFF study, in 2026, the average annual cost of family medical insurance through an employer increased by 5% to $28,303. Of this amount, employees pay an average of $7,471, which is 9% more than the previous year.

Significant administrative burdens are created by pre-authorizations for treatment with insurance companies. According to the American Medical Association, one doctor processes about 40 such requests per week on average, requiring approximately 13 hours of work time from doctors and staff. The American Hospital Association estimates total administrative expenses of the U.S. healthcare system to be 25–35% of its total costs.

Google aims to reduce part of this burden through automatic document preparation and insurance condition checks. For example, Gemini can match medical history with insurance company rules, identify missing information, and prepare a document package for review. If an insurer denies authorization, the system can help draft an appeal with medical data and coverage conditions.

For insurance organizations, separate scenarios are provided for application verification, coverage rule analysis, and patient expense calculation. However, final decisions related to treatment and insurance denials should be made by authorized specialists, not automatically by the model itself.

Can hospitals really reduce costs

John D'Angelo, head of Northwell Health, believes that fragmented medical data leads to additional examinations and administrative expenses. According to him, doctors often have to search for information on previously conducted procedures, and in the absence of available information, they may order repeat tests.

Northwell hopes that automating information retrieval will reduce paperwork and allow more time for patients. The company is also considering using AI for preparing medical records and pre-checking insurance coverage before appointments. However, these are expected benefits rather than results from a published study on the economic efficiency of Gemini for Healthcare.

The platform supports the integration of external medical systems and the use of different AI models under unified organizational management. Google states that its architecture is designed to comply with HIPAA requirements and includes access control to confidential data. However, medical institutions need to independently verify the accuracy of AI results and compliance with regulatory requirements.

At the time of launch, Google Cloud did not disclose the timeline for Gemini for Healthcare's transition to full commercial availability and did not provide data on confirmed reductions in medical service costs. The primary goal of the first phase will be to test how well AI can speed up work with documents and insurance systems without compromising the quality of medical care.