Joint Commission’s New Framework for Responsible AI

[00:00:00] All right, welcome back to the DocBuddy Journal, the podcast from your friends at DocBuddy, where we cut through the noise of healthcare tech and help clinicians, administrators, and health systems navigate what’s real, what’s hype, and what actually matters at the bedside and in the boardroom. I’m your host, Eric, and today we’re talking about a headline that landed June 1st, 2026.

And honestly, it’s one of the more significant announcements we’ve seen in the healthcare AI space in quite a while. This news is that the Joint Commission, the nation’s oldest and largest healthcare accreditation body, which evaluates more than twenty-three thousand healthcare orgs across the United States, has officially launched what it’s calling the Responsible Use of AI in Healthcare certification, or RUAIH for those of us who love a good acronym.

Here at DocBuddy, we’ve been watching the AI in healthcare space closely, and we have thoughts. Stick with us.

[00:01:00] 

Let’s start with the basics. What exactly is this thing? The RUAIH certification is a voluntary program, and I want to underscore that word voluntary because we’ll come back to what that means, uh, which is designed to recognize US hospitals, critical access hospitals, and health systems that can demonstrate they have the governance, safeguards, monitoring processes, and education in place to use AI responsibly.

Now, here’s something important to understand right out of the gate. This certification does not validate or certify individual AI tools or products. The Joint Commission is not stamping a particular algorithm or AI vendor with a seal of approval. Instead, what they’re certifying is whether the organization deploying the AI has done the work to deploy it safely.

That’s a crucial distinction. An AI tool can be technically brilliant and still cause harm if the organization using it hasn’t thought through governance, bias, data quality, and patient safety [00:02:00] implications. The Joint Commission gets that. The certification standards are organized around five major pillars.

The first being governance. Who’s in charge? Who’s accountable? The next, effective data management. Is the data feeding your AI systems accurate, representative, and protected? Three, risk and bias reduction. Are you actively identifying and mitigating the ways AI can fail certain patient populations? The fourth, monitoring, evaluating, and validating safety performance.

This isn’t a set it and forget it situation. Are you watching what your AI is doing after deployment? And five, finally, transparency, education, and training. Do your clinicians and staff actually understand the AI tools they’re using? Five pillars. That’s the framework And this framework didn’t come out of nowhere.

The Joint Commission has been building towards this for over a year. Back in 2025, the organization released initial guidance to [00:03:00] help US health systems implement AI safely and effectively. That guidance wasn’t developed in a vacuum. It came after the Joint Commission convened more than 20 coalitions and groups with expertise in both healthcare and technology.

So that’s a lot of stakeholder input, and that’s the kind of deliberate, measured process you want to see when you’re setting standards that will affect patient care. The organization that has collaborated closely with Joint Commission on this work is the Coalition for Health AI, or CHAI. Their CEO, Dr.

Brian Anderson, released a statement saying, and I, I’m paraphrasing here, that CHAI’s recently published governance playbooks and the Joint Commission certification are tightly aligned on the need for responsible and transparent AI in healthcare, and that this alignment will help reduce confusion while accelerating responsible AI adoption.

Now, reduce confusion, that’s a phrase worth sitting with because right now in healthcare, there is a lot of confusion. Every vendor has a different safety story. Every [00:04:00] health system is figuring it out on their own. A shared framework or a common language, that’s gonna be genuinely valuable as we navigate these what are still new AI headwaters

Why now? Well, the numbers tell the story. Here’s the context that makes this announcement urgent. According to a March 2026 report from the AMA, and that’s the American Medical Association, eighty-one percent of surveyed physicians, so that’s obviously more than eight in ten, are now using AI in clinical practice.

Compare that to just sixty-six percent in 2024 and only thirty-eight percent in 2023. That’s not gradual adoption. That’s a full-out sprint. The Joint Commission’s president and CEO, this is Jonathan Perlin, uh, MD, PhD, cited the same figure when he said there is a, quote, “Fast-growing need for universal standards for implementing this transformational technology in responsible ways,” end quote.

He went on [00:05:00] to say AI has the potential to unlock discoveries and improve quality, safety, and operating efficiency. With this new certification, Joint Commission is providing healthcare orgs with a blueprint for safety– safely and appropriately using AI. And that’s an enthusiastic statement. Uh, and at DocBuddy, we would agree with the sentiment.

AI does have enormous potential in healthcare, and we see it every day in our own work and on our own solutions, such as Enterprise AI and AI Notes, which you can learn more about at docbuddy.com. Uh, but here’s the thing, and this is a point we wanna make very clearly today. Potential is not performance.

The fact that eighty-one percent of physicians are using AI doesn’t mean all of those deployments are safe, equitable, or effective. It means adoption has outpaced oversight, at least to some degree, and that’s exactly the gap that this certif-certification is trying to close So let’s bring in some voices from healthcare organizations that are living this reality.

And unfortunately, we weren’t able to get their [00:06:00] actual voices, so I’m gonna do my best, uh, set of impressions here. Uh, this first, uh, piece of feedback is from Aaron Mary. Aaron is the Executive Vice President and Chief Digital Information Officer at Baptist Health. Uh, and this Baptist Health is located in Jacksonville, Florida.

Not to be confused with Baptist Health South Florida down by me in Miami. Uh, Aaron offered this perspective in support of the certification. He described seeing AI make processes more efficient and staff more productive every day. He emphasized that it’s enabling nurses and doctors to spend more time doing what they love, which is delivering quality care to patients and their families.

Uh, but he also emphasized, and, and this is important, that it is critical that hospitals have a framework to follow for this emerging and evolving tech. He called this new certification, quote, “long-awaited,” end quote, by Baptist Health and many other orgs across the industry, particularly as AI tools become increasingly embedded in clinical, operational, administrative, and care [00:07:00] support workflows.

Long-awaited? Well, that’s telling. Health systems have been deploying AI tools and holding their breath, hoping that governance frameworks would catch up, and this announcement is the field saying, “The guardrails have arrived. Let’s be sure that we use them.” So now for our take here. It’s gonna center around measure twice, cut once, and here’s where we get off of this press release, press release and share what we actually think here at DocBuddy.

We are genuinely excited about the certification, not because it’s a bureaucratic checkbox, quite the opposite. And any longtime listener of the show will know we’re not here, and I’m not personally here to hope for regulations, to hope for oversight, to hope that our government, uh, gets more involved in the practice of medicine than it already has.

Um, we’re excited because it reflects a little bit of a maturing understanding of what AI and healthcare actually requires. AI is not a panacea. How many times have you [00:08:00] heard me say that? It’s a tool, a powerful, promising, increasingly capable tool, and of course, I’m speaking broadly here. Um, I’m gonna shout out one more time that if you’d like to learn more about our EHR-integrated AI-powered solutions like Enterprise AI and AI Notes, you can learn all about them on docbuddy.com.

Of course, I’m speaking more broadly about healthcare, uh, and health tech and AI and healthcare, uh, broadly here, generally speaking. So this tool that’s powerful, promising, and increasingly capable, but still a tool nonetheless, can fail. These tools can hallucinate, discriminate, and mislead when it’s deployed carelessly.

The healthcare industry, which more than perhaps any other, cannot afford to learn that lesson the hard way. And that brings us to the old carpenter’s principle, measure twice and cut once. Think before you act. Validate before you deploy. Understand what you’re doing before you do it. That principle has never been more important than it is in [00:09:00] healthcare AI right now.

And part of the difficulty in all of this is that healthcare and AI can mean many, many different things and address many, many different types of workflows. And this is exactly, uh, what the Joint Commission’s RUAIH framework is asking healthcare systems to do. It’s not asking you to slow down innovation.

It’s asking you to slow down enough to make sure you’ve done the governance work, to make sure that you’ve done the bias testing, to be sure you’ve done the monitoring infrastructure and the staff training groundwork before AI touches a patient encounter. Think about those five pillars one more time: government, data management, risk and bias reduction, ongoing monitoring, transparency and training.

None of these are exotic concepts. They are the basics of responsible technology deployment. The problem is that in the rush to adopt AI, which is in some part driven by vendor pressure, we see a lot of things on trade show floors, even thinking back to ASCA, a lot of things are getting called AI that maybe aren’t actually [00:10:00] AI.

They’ve got this vendor pressure. You have competitive anxiety, keeping up with the Joneses, and then more excitingly, genuine enthusiasm for the technology’s potential. Many healthcare orgs have either skipped or looked past these, these steps, which equates to them cutting before measuring carefully. So this Joint Commission RUAIH framework is providing a structured prompt to just stop and measure.

I think we’re bringing it home here. Let’s talk about what this means practically speaking. First, you do not need to be accredited by the Joint Commission to purf– to pursue this cert. It’s open to any healthcare org in the US, and that matters because it means community hospitals, rural critical access facilities, and smaller health systems can pursue this framework too.

This isn’t just for the big academic medical centers. Second, because this is voluntary right now, there’s a real question about uptake. Will organizations that [00:11:00] need this most, the ones deploying AI without adequate governance, actually pursue it? Or will it primarily attract organizations that are already doing the right things and want to demonstrate it?

That’s a legitimate question. And in, uh, in stats, if you think back to your business stats class, you’ll know this is something called response bias. Uh, plus, it’s one of the healthcare– one that the healthcare industry will need to answer together over the next few years. Uh, our hope is that the field embraces this not as a marketing credential but as a genuine operational standard.

Third, this cert doesn’t evaluate AI vendors. If you’re a health system evaluating an AI product, the RUAIH, and man, that’s a tough acronym, uh, gives you a framework for assessing your organization’s readiness, but it doesn’t do the vendor due diligence for you. That work still has to happen. You still need to interrogate the training data, understand the model’s limitations, pressure-test it in your patient population, and then [00:12:00] again, be measuring twice and cutting once.

AI vendors will want to tell you that their tool is safe and effective. Your job is to verify that and to verify that your organization is ready to deploy it safely. That’s two different questions, and they both need answers. DocBuddy’s AI solutions, of course, uh, will, uh, be safe and effective for your organization.

Would be happy to show you why. Docbuddy.com, uh, slash contact to request a demo. Sorry, couldn’t help myself but plug that one there. So let’s zoom out for just a moment. We’re gonna take a look at the bigger picture. The Joint Commission, which has been around since 1951, has spent more than seventy years setting the standards for healthcare quality and patient safety in this country.

And of course, there are other, uh, accreditation bodies. Uh, we’re singling out the Joint Commission here because we’re talking about their voluntary AI certification, of course. We’re not playing favorites. We’re simply providing you some news. Um, and when they [00:13:00] move into a new domain, it matters, and it signals that the domain has reached a level of maturity and risk that warrants universal standards.

That they are now issuing a cert, certification, of course, specifically for AI governance is in itself a statement. AI is no longer an experiment. It’s infrastructure, and infrastructure needs standards. The healthcare IT news coverage of the story noted that RUAIH certification is meant to ensure AI’s benefits outweigh its risks, including privacy and security concerns, data inaccuracies, and the lack of transparency in many AI decision-making processes.

These aren’t hypothetical risks. They are documented, recurring challenges that show up in deployment after deployment. As a patient and as a healthcare stakeholder myself and, uh, on, on speaking on your behalf as well, these are things you want your healthcare organizations, your provider organizations to be thinking about and acting upon.

And the fact that a body as established and credentialed as the Joint Commission is now [00:14:00] formalized around addressing those risks should give every health system a moment of honest reflection. Have we done this work? Do we have governance in place? Are we monitoring our AI tools after they go live? Are our clinicians trained to understand what the AI can and cannot do?

If the answer to any of those questions is, “I’m not sure,” the RUAIH framework is a roadmap you need to be thinking about

So let’s wrap it up. I think we’ve spent, uh, a fair amount of time here. It’s been fantastic. Here at DocBuddy, of course, our mission is to give providers and their care teams more time for life, uh, for patients, and for the people delivering it. And we believe tools like AI thoughtfully deployed can and will be transformative.

But thoughtfully deployed is doing a lot of work in that sentence. The Joint Commission’s new RUAIH cert is the most significant structural step the healthcare industry has taken to define what thoughtfully deployed actually looks like. We’re [00:15:00] encouraged, we’re watching, and we hope that the health systems across the country, large and small, will engage with this framework not as a compliance exercise, but as a genuine commitment to their patients.

Measure twice, cut once. That’s the DocBuddy Journal for this week. If you found the show interesting or valuable even, please subscribe, share it with a colleague in healthcare, and visit us at docbuddy.com for more resources on navigating AI in clinical and operational settings responsibly. Until next time, thank you for listening on behalf of the entire DocBuddy team.

Take care of your patients, take care of yourselves, and we’ll talk to you again soon.