Artificial intelligence is poised to revolutionize healthcare delivery, and Google is taking significant steps to collaborate with healthcare providers to gather evidence on its most effective applications, as highlighted in a recent feature by Modern Healthcare.
Michael Howell, M.D., the chief health officer at Google, emphasized the company's goal of creating a “safe testing harness” for AI in the medical field. This involves conducting research to identify the best practices for integrating artificial intelligence into healthcare. Like other major tech players such as Apple, Amazon, and Microsoft, Google is forming partnerships with health systems to drive the adoption of AI technologies that enhance experiences for both providers and patients.
After the surge of interest in large language models (LLMs) following the launch of ChatGPT by OpenAI, Google began to delve into the clinical capabilities of conversational AI. The company unveiled the Articulate Medical Intelligence Explorer (AMIE), a research-driven AI tool tailored for diagnostic reasoning and patient interactions.
In a blog post, research leads Alan Karthikesalingam and Vivek Natarajan noted that AMIE was rigorously trained and assessed against criteria reflective of quality clinical consultations, taking into account perspectives from both healthcare professionals and patients.
Recently, Google has expanded its collaborations with a variety of entities in the healthcare sector, including digital health firms and traditional health systems. A notable partnership announced in February 2026 with Included Health—a telemedicine provider—aims to conduct a nationwide randomized controlled trial investigating the role of conversational AI in enhancing virtual care processes.
“This research will further build on our foundational work involving AI for diagnostic reasoning, personal health insights, and navigating health information,” explained research leads Mike Schaekermann and Cameron Chen in a blog entry.
This study builds on Google's earlier investigations into AI's potential to improve virtual care delivery, including efforts that showed AI could succeed in passing medical licensing tests designed for human doctors, according to Dr. Howell. In a practical test, simulated patients with conditions like heart failure and breast cancer evaluated the capabilities of this conversational AI.
Further collaboration with Harvard involved 120 real patients interacting with the AI under physician supervision, with positive outcomes reported by Howell to Modern Healthcare.
In addition to partnerships with health organizations, Google has joined forces with the Mayo Clinic in Rochester, Minnesota. Dr. Howell pointed out that Google's research initiatives and efforts through Google DeepMind have bolstered Mayo’s healthcare advancements. In the realm of radiotherapy, for instance, research has demonstrated that AI tools developed by Google can significantly improve efficiency and standardization for patients with head and neck cancers.
A 2023 press release from Mayo Clinic stated that AI-generated contours for radiotherapy planning were deemed clinically viable with minimal revisions 90% of the time, compared to just 53% for traditional methods. This advancement effectively reduced the overall time for contouring and related processes by 76%.
The partnership with Google has broader implications, grounded in a 10-year agreement established in 2019 aimed at promoting ethical clinical data usage. As detailed in a National Academy of Medicine case study, this collaboration has granted Mayo access to advanced AI tools, security expertise, and engineering support, ultimately aiding in the development of the Mayo Clinic Cloud and Platform, which allows for secure sharing of de-identified patient data.
“This partnership illustrates how health systems can collaborate with technology companies to foster knowledge generation while effectively addressing privacy and cybersecurity issues,” states the case study abstract, highlighting the dual goals of facilitating data collaboration and offsetting costs associated with algorithmic development.
Google's contributions also extend to creating “pathway assistants” that utilize Google Cloud and Gemini models to streamline access to essential clinical information, aiding healthcare providers at critical moments by transforming previously inaccessible knowledge housed in manuals and digital files.
This movement towards enhanced healthcare through AI isn’t limited to Google alone. Other tech giants are similarly investing in AI to improve clinical care. Recently, Microsoft and Mayo Clinic announced initiatives to merge Microsoft’s AI capabilities with Mayo’s clinical insights, aiming to develop a specialized AI model for healthcare applications that will be accessible through Azure Foundry APIs.
Amazon has also entered the healthcare AI sphere, unveiling Amazon Connect Health—a sophisticated AI solution for healthcare that, among other functionalities, improves patient identity verification and scheduling tasks, saving one health system over 630 hours weekly.
As the industry explores these innovative opportunities, it remains crucial to maintain a careful balance between advancement and safety. In 2021, Mayo Clinic established its Health Data and Technology Advisory (DaTA) Board to gauge the impact of AI and health technology from diverse patient perspectives, as noted in the National Academy of Medicine’s analysis.
Google, too, is advancing cautiously, focusing on a gradual and methodical approach to AI integration in healthcare settings, continually accumulating data about its efficacy and impact.
“We believe it is vital to build an evidence base to clarify where AI is effective and where it may fall short,” Dr. Howell stated in the discussion with Modern Healthcare.
