The Coalition for Health AI (CHAI) on Wednesday released a series of detailed playbooks aimed at providing health systems with practical guidance and baseline governance controls to safely, transparently, and consistently implement artificial intelligence amid rapid adoption.
While these handbooks touch on specific controls to enable responsible use of AI across healthcare organizations, the nonprofit says each is intended to be “interpreted within the specific context” of individual organizations.
According to CHAI, each handbook includes recommended implementation guidance, tools, resources, and examples for healthcare organizations to integrate into their existing processes.
The eight areas covered by this framework are:
- Organizational AI policy
- organizational structure
- organizational resources
- Responsible AI lifecycle management
- Risk and impact assessment
- Responsible data management and use
- Third party management
- Education, training and feedback
The CHAI handbook was developed in collaboration with more than 100 healthcare organizations through workshops and multiple calls for feedback.
The playbook can be accessed on the CHAI website.
CHAI CEO Brian Anderson, MD, said in a statement that the organization is hearing a growing need for “practical, consistent, and comprehensive” guidance on AI adoption.
“Our working group set out to address this challenge, and we are encouraged by the amount of collaboration that has occurred from across the industry to make today’s handbook a reality,” Anderson said. “These efforts not only help define responsible AI, but also aim to make AI available to healthcare delivery organizations across the country, with the goal of turning AI innovation into quality care for all patients, regardless of resource level.”
The governance strategy follows an earlier document published last autumn.
In September, CHAI partnered with the Joint Commission on the Responsible Use of AI in Health Care (RUAIH) guidance document to outline key principles for organizational governance of the use of AI in health systems. This follows the announcement of a partnership between the two organizations in June.
The Joint Commission plans to offer a voluntary AI certification program this year, and the CHAI Handbook provides a framework for achieving certification.
Mercy Health Responsible AI Program Director Taylor Rhodes said in a statement that resource structure is “one of the most important challenges” facing healthcare AI today, which Rhodes said is “transforming good intentions into managed, measurable, and sustainable practices.”
“These provide health systems with a common operating language for responsible AI while allowing each organization to adapt governance to its unique mission, workflow, maturity, and risk tolerance,” Rose said. “This work reflects where healthcare AI governance needs to go next: not just setting expectations, but building the reproducible evidence, ownership, and oversight needed to deploy AI securely, transparently, and with trust.”
