Blab AI
UCLA Health Launches INOVAi Center to Test AI Safety and Effectiveness in Clinical Settings
On June 11 2026, UCLA Health announced the creation of the Innovations and Outcomes Validation of AI (INOVAi) Center, a research hub that will evaluate artificial intelligence (AI) tools in real‑world healthcare settings.
The center is one of the first national programs dedicated specifically to the evaluation and implementation science of AI in medicine. It will track AI systems from early usability and workflow testing through prospective clinical trials and pragmatic implementation studies, with the aim of bridging the gap between AI development and large‑scale deployment by refining models through rigorous, evidence‑based evaluation.
Johnese Spisso, president of UCLA Health and CEO of the UCLA Hospital System, said the center will address a critical gap: “Knowing whether these tools are safe, effective and useful in real‑world clinical practice.” He added that rigorous evaluation across patient‑care settings can help ensure AI strengthens clinical practice and works reliably for all patients.
INOVAi is managed through the UCLA Department of Medicine and operates in alignment with the UCLA Center for AI & SMART Health. Steven Dubinett, dean of the David Geffen School of Medicine at UCLA, described INOVAi as a Center of Excellence in AI evaluation and implementation science. He noted that the center will strengthen a coordinated ecosystem for AI innovation that is rigorous, ethical, evidence‑based and focused on improving health.
Katherine Andriole, associate dean for health AI strategy and innovation at the medical school and director of the UCLA Center for AI & SMART Health, emphasized the responsibility to approach AI thoughtfully. She said the clinical validation work performed by INOVAi is essential as AI continues to transform care delivery.
INOVAi’s leadership includes Dr. Paul Lukac, UCLA Health’s chief AI officer, and Dr. John Mafi, associate professor of medicine. The two co‑directors led a randomized trial that found AI scribes significantly reduced the time physicians spent writing clinical notes. The study also reported improvements in physician cognitive load, work exhaustion, and overall well‑being, as well as enhanced patient engagement through greater connectivity.
Lukac noted that the results of the research on AI’s capacity to improve clinical workflow and patient care are promising but that further work is needed. He said the new program has the potential to hone the discipline required to build a shared language for what counts as evidence.
INOVAi’s focus on the full AI lifecycle reflects the growing need for systematic evaluation as AI tools such as computerized scribes, diagnostic assistants, and imaging interpreters become more common across U.S. clinical settings. By conducting usability studies, feasibility assessments, and pragmatic trials, the center aims to provide data that can inform clinicians, regulators, and developers.
The launch of INOVAi follows UCLA Health’s broader strategy to promote responsible health AI. The institution has already established the UCLA Center for AI & SMART Health and has integrated AI research into its clinical operations. The new center is intended to complement these efforts by providing a structured framework for testing and implementing AI solutions.
While the center’s first projects focus on AI scribes and workflow integration, its mandate includes evaluating a wide range of AI applications, from diagnostic decision support to predictive analytics. The center will also collaborate with industry partners, academic researchers, and patient communities to ensure that safety, effectiveness, and equity remain central to AI deployment.
In summary, the INOVAi Center represents UCLA Health’s commitment to evidence‑based AI adoption in medicine. The center’s work will generate data on safety, effectiveness, and real‑world impact, helping to guide clinicians, developers, and policymakers as AI tools become increasingly integrated into patient care.
The center’s initial studies, including the AI scribe trial, are already underway. Future phases will expand to additional AI modalities and larger patient populations, with the aim of producing robust, generalizable evidence that can inform broader healthcare AI implementation.
The center is one of the first national programs dedicated specifically to the evaluation and implementation science of AI in medicine. It will track AI systems from early usability and workflow testing through prospective clinical trials and pragmatic implementation studies, with the aim of bridging the gap between AI development and large‑scale deployment by refining models through rigorous, evidence‑based evaluation.
Johnese Spisso, president of UCLA Health and CEO of the UCLA Hospital System, said the center will address a critical gap: “Knowing whether these tools are safe, effective and useful in real‑world clinical practice.” He added that rigorous evaluation across patient‑care settings can help ensure AI strengthens clinical practice and works reliably for all patients.
INOVAi is managed through the UCLA Department of Medicine and operates in alignment with the UCLA Center for AI & SMART Health. Steven Dubinett, dean of the David Geffen School of Medicine at UCLA, described INOVAi as a Center of Excellence in AI evaluation and implementation science. He noted that the center will strengthen a coordinated ecosystem for AI innovation that is rigorous, ethical, evidence‑based and focused on improving health.
Katherine Andriole, associate dean for health AI strategy and innovation at the medical school and director of the UCLA Center for AI & SMART Health, emphasized the responsibility to approach AI thoughtfully. She said the clinical validation work performed by INOVAi is essential as AI continues to transform care delivery.
INOVAi’s leadership includes Dr. Paul Lukac, UCLA Health’s chief AI officer, and Dr. John Mafi, associate professor of medicine. The two co‑directors led a randomized trial that found AI scribes significantly reduced the time physicians spent writing clinical notes. The study also reported improvements in physician cognitive load, work exhaustion, and overall well‑being, as well as enhanced patient engagement through greater connectivity.
Lukac noted that the results of the research on AI’s capacity to improve clinical workflow and patient care are promising but that further work is needed. He said the new program has the potential to hone the discipline required to build a shared language for what counts as evidence.
INOVAi’s focus on the full AI lifecycle reflects the growing need for systematic evaluation as AI tools such as computerized scribes, diagnostic assistants, and imaging interpreters become more common across U.S. clinical settings. By conducting usability studies, feasibility assessments, and pragmatic trials, the center aims to provide data that can inform clinicians, regulators, and developers.
The launch of INOVAi follows UCLA Health’s broader strategy to promote responsible health AI. The institution has already established the UCLA Center for AI & SMART Health and has integrated AI research into its clinical operations. The new center is intended to complement these efforts by providing a structured framework for testing and implementing AI solutions.
While the center’s first projects focus on AI scribes and workflow integration, its mandate includes evaluating a wide range of AI applications, from diagnostic decision support to predictive analytics. The center will also collaborate with industry partners, academic researchers, and patient communities to ensure that safety, effectiveness, and equity remain central to AI deployment.
In summary, the INOVAi Center represents UCLA Health’s commitment to evidence‑based AI adoption in medicine. The center’s work will generate data on safety, effectiveness, and real‑world impact, helping to guide clinicians, developers, and policymakers as AI tools become increasingly integrated into patient care.
The center’s initial studies, including the AI scribe trial, are already underway. Future phases will expand to additional AI modalities and larger patient populations, with the aim of producing robust, generalizable evidence that can inform broader healthcare AI implementation.
