Health systems are updating their AI governance policies to keep pace with rapid advances in the technology and growing interest among frontline teams.
Cedars-Sinai in Los Angeles and Cleveland Clinic shared how their internal frameworks for managing AI have evolved since initial implementation — and how they are keeping governance flexible as tools mature.
Cedars-Sinai’s AI efforts started with a focus on safety, alignment with strategic goals and performance monitoring — what Mouneer Odeh, the system’s inaugural chief data and artificial intelligence officer, calls the “AI enablement” phase. Now, the organization is entering a new stage centered on scaling and cultural transformation.
“We are building on strong momentum to deploy many more use cases, with more focus on the cultural implications as we democratize access to these powerful AI tools,” Mr. Odeh told Becker’s. “We want to foster a culture that embraces frontline innovation and encourages our team members to proactively acquire new skills.”
At Cleveland Clinic, early AI governance focused on building a multidisciplinary team to guide decision-making. That team included representatives from clinical care, legal, compliance, cybersecurity, bioethics, and data and analytics.
“AI governance required a multidisciplinary approach,” Jennifer Owens, senior AI program administrator at Cleveland Clinic, told Becker’s. “With the establishment of our AI Task Force, we are now able to focus on solidifying ethical principles by which AI proposals are evaluated.”
The policy now includes guidance on enterprise-approved tools and is shaped by new developments in the field, including updates from national organizations such as the Coalition for Health AI.
Both health systems say flexibility is essential in their governance strategies.
Cedars-Sinai has developed internal tools and new collaboration methods to monitor the performance and safety of AI models, particularly when vendors are unable to validate results using the system’s own data.
“AI is evolving rapidly, and many companies developing AI solutions don’t have robust tools to validate or monitor performance on our data,” Mr. Odeh said. “This led us to develop new tools and collaboration methods to ensure the safety and effectiveness of the models we deploy.”
At Cleveland Clinic, the governance policy is formally reviewed on an annual basis to account for changing risks and product capabilities.
“As existing products gain new features, it’s important that we remain flexible to adapt our policies to meet the current needs,” Ms. Owens said.
As AI tools continue to evolve, leaders at both health systems say that dynamic governance — built with flexibility and cross-disciplinary input — will remain key to balancing innovation with safety.
