As AI becomes common across healthcare technology, the voluntary certification gives hospitals a way to demonstrate they have processes in place to govern its use.
By Alyx Arnett
The Joint Commission has launched its Responsible Use of AI in Healthcare certification, a voluntary program for healthcare organizations that want to demonstrate they have processes in place to govern AI use.
The certification builds on AI guidance released by Joint Commission and the Coalition for Health AI in 2025 and governance playbooks released in May. The certification gives organizations a way to demonstrate they have implemented those recommendations.
William Walders, MHA, MBA, chief digital and information officer at Joint Commission, says the certification comes as AI capabilities become more common in medical devices and software. That makes AI oversight a patient safety issue for healthcare technology management (HTM) teams, he says, as they evaluate and support many of those technologies.
“The most important technology in the health system is that infusion pump,” Walders says. “It’s not the ERP. It’s not email. Why? Because it’s keeping patients alive.”
What the Certification Looks For
To earn certification, organizations must demonstrate processes in five areas: governance, data management, risk and bias reduction, monitoring and validation, and transparency, education, and training.
Ken Grubbs, DNP, MBA, RN, chief nurse executive at Joint Commission, says the certification standards are intended to be flexible rather than prescriptive so that organizations can adapt them to their own AI use cases. For organizations considering certification, he recommends starting with the AI guidance and governance playbooks, then comparing existing technology evaluation, training, safety, and reporting processes against the certification requirements. “Look at what you already have in place,” he says. “Simply do a gap assessment and determine if there are any gaps.”
Walders says many organizations may find much of the needed structure already exists. Rather than creating entirely new evaluation processes, he says HTM teams can build on the questions they already ask when assessing new technologies.
“You understand the modalities built into a CT; you understand the protocol built into an IV pump. Now ask one level deeper: Do I know if it’s AI doing that or if it’s the way the software is built?” Walders says.
HTM’s Role in AI-Enabled Devices
Walders says HTM teams should apply the same rigor to AI-enabled systems that they do to other high-risk technologies. That includes validating how the technology works, ensuring staff understand how to use it, monitoring its performance after deployment, and continuing to follow existing reporting processes.
“Don’t set it and forget it,” Walders says. “You have to monitor what it looks like post-deployment. What’s the tool doing? Where is it getting [information] from? Is there a human in the loop, and is the human actually in the loop?”
Grubbs says deploying AI also requires collaboration between HTM, IT, and clinical leaders, noting that responsibility for AI extends beyond a single department.
“Rolling out AI can’t be done in a vacuum,” Grubbs says. “It takes all members of the team, as well as the clinical leaders.”
When evaluating AI-enabled products, Walders recommends using Joint Commission’s AI guidance as a starting point. HTM teams should understand how the technology is governed and trained, what data it uses, how users will be trained, and how performance will be monitored over time.
“Those are the same questions I’d ask,” Walders says. “We’re guiding you to ask those questions of yourself. Naturally, your third-party partner should do the same.”
An Ongoing Program
Grubbs says the certification is not meant to be a one-time exercise. The Joint Commission expects the program to evolve as healthcare AI use changes, and certified organizations are expected to continue embedding the requirements into safety and quality processes.
“This certification isn’t about obtaining the certification and saying, ‘OK, we’ve obtained the certification, and we’re done,’” Grubbs says. “It’s about continual implementation.”
Walders says the certification demonstrates that organizations have a structured approach to AI governance.
“It signals that you’re looking at it, that there’s trust, and we’re reducing risk,” he says. “We’re not just adopting AI to be first. [It’s] responsible adoption.”
The certification carries a fee, and organizations interested in applying can contact Joint Commission or, if already accredited, reach out to their Joint Commission account executive for more information.
ID 419780622 © BiancoBlue | Dreamstime.com
Alyx Arnett is chief editor of 24×7 Magazine. Questions or comments? Email [email protected].
