Blogs

As healthcare delivery organizations move from experimenting with AI to deploying it across clinical settings, governance has become a leading challenge. On May 27, 2026, CHAI released its series of in-depth governance playbooks to help its members and the broader ecosystem manage this process. The playbooks were developed by workshops and workgroups that included 150+ health AI leaders, and directly informed Joint Commission's Responsible Use of AI in Healthcare certification program.

Christopher Lee, MBA, BSN, RN-BC, Clinical Nurse and Interim Practice Transformation & Innovation Specialist at UCLA Health, participated in the development process of these playbooks, bringing both a frontline clinical and nursing professional perspective on governance. We spoke with Christopher about what healthcare organizations can learn from the collaborative process behind the playbooks, how the guidance can translate across organizations with vastly different resources, and how it can help healthcare delivery organizations prepare for emerging expectations around responsible AI.

What did developing the Governance Playbooks alongside stakeholders from across healthcare bring to the process that you couldn't get from a framework developed in isolation?

What stood out to me about the CHAI process was the diversity of perspectives, experiences and skill sets involved. There were frontline clinical voices alongside people working in research, policymaking, government relations and other areas of healthcare. That diversity helped connect what clinicians experience on the ground with the structures, processes and guidance needed to govern AI effectively. There was also a real sense of trust and mutual respect. Everyone came in with a shared goal of responsible AI implementation, which allowed people to openly challenge ideas and identify areas that could be improved.

From a nursing perspective, one tangible example was making sure nurses were explicitly named when discussing who should be represented in AI governance. Language is powerful. If you don't identify who needs to be in the room, you can't assume their perspective will be represented.

How did participants' real-world experience help shape the guidance in the Playbooks?

Healthcare is inherently multidisciplinary, and the development process reflected that. For instance, a nurse may recognize an impact on patient care or workflow that someone in another role doesn't immediately see. And a physician may identify something affecting clinical decision-making, while an operational leader brings another important perspective.

Those viewpoints matter because AI doesn't operate in isolation. I'm seeing more and more that the strongest adoption and implementation strategies bring together stakeholders across nursing, medicine, care coordination, operations, informatics, and other relevant areas. This may take the form of focus groups, advisory councils, or other structured opportunities for end users to evaluate and provide feedback on technologies such as AI-enabled EHR tools or risk-detection systems.

That kind of involvement is important throughout the AI lifecycle, from evaluation and implementation through ongoing monitoring and decisions about whether a technology should continue to be used.

What makes the Governance Playbooks useful for organizations that may not have extensive AI governance resources internally?

I see the playbooks as practical, adaptable and evergreen resources. One of their greatest strengths is that they can help organizations incorporate responsible AI principles regardless of their size, resources, or care setting.

A large academic health system may have significant internal resources to devote to governance, while a rural or critical access hospital may not have the capacity to develop an AI governance framework from scratch. The playbooks can help level the playing field by providing a common foundation that organizations can adapt to their own needs and cultures.

We started from shared principles of responsible AI, including transparency, safety, security and equity, and considered what organizations need to put those principles into practice: clear structures, process and workflow owners, stakeholder involvement and standardized approaches to implementation and oversight.

Joint Commission's Responsible Use of AI in Healthcare certification addresses areas including governance, data management, patient safety, quality monitoring and education. Having worked through CHAI's Governance Playbooks, how do you see this guidance helping organizations prepare for those kinds of expectations?

What gives me confidence in the playbooks is the consensus model CHAI used to develop them. There were open feedback sessions where we could challenge what had been put on paper, and the guidance benefited from perspectives across the healthcare ecosystem.

I think the playbooks set a strong foundation for where organizations should start across the AI lifecycle. That foundation is important as healthcare organizations face more formal expectations around how AI is governed, monitored and evaluated.

AI is evolving incredibly quickly, so no framework can be static. But having a structured starting point that has been challenged by people with different roles and experiences can help organizations think systematically about responsible AI rather than having to build their approach from the ground up.

Where do you think healthcare AI governance still needs to evolve?

AI capabilities are changing at an accelerated rate, so governance needs to evolve alongside the technology. We need to continue thinking about how we anticipate and mitigate risk, particularly as AI becomes more powerful and more deeply embedded in healthcare workflows.

That includes maintaining meaningful human oversight. AI outputs depend on the context and data they're grounded in, and we shouldn't assume that a technology will always produce the right result. Governance therefore can't end at deployment. Organizations need to consider how technologies are monitored and held accountable throughout their lifecycle.

As organizations put these governance structures into practice, what do you hope they keep at the center?

Ultimately, I hope we're using AI to create more space for humanism in healthcare: more time for compassion, active listening and meaningful interactions with patients.

We need to be careful that efficiencies created by AI don't simply translate into greater workload expectations. We should also ask how these technologies can give clinicians time back for the fundamentally human aspects of care.

And I would encourage more nurses to become part of these conversations. Nurses represent one of the largest segments of the healthcare workforce, and our perspective is important when decisions are being made about technologies that will increasingly shape how we care for patients.