Joint Commission CEO Breaks Down New AI Certification
Based on facts, either observed and verified firsthand by the reporter, or reported and verified from knowledgeable sources.
Newsweek AI is in beta. Translations may contain inaccuracies—please refer to the original content.
This is a preview of Access Health—Tap here to get this newsletter delivered straight to your inbox.
I've settled on my word of the week: transparency. Over the last seven days, my schedule has looked like a patchwork quilt, spanning several distinct sectors of the health care industry. But from the government's roundtable on prior authorization reform to my conversation with Joint Commission CEO Dr. Jonathan Perlin (read on for more on both of those), transparency has emerged as a common thread.
Health systems and insurance companies must be transparent with one another; health systems and insurance companies must be transparent with patients. And, of course, health care companies must be transparent with their employees and customers—especially when deploying AI.
Monday through Wednesday, many of my colleagues were in sunny Sonoma, California, for Newsweek's cross-industry AI Impact Summit. Senior Reporters Lauren Giella and Katherine Fung kept eyes on the health care side of the agenda, which included speakers from Kaiser Permanente, Hospital for Special Surgery, AdventHealth and UMass Medical.
"Transparency was a big theme for AI adoption in health care—not only when dealing with medical records and personal data, but also for why and how organizations are implementing automation tools," Giella wrote.
Dr. Allen Chang, ACMIO at UMass Medical, warned the audience not to neglect employees' concerns about losing jobs to AI, nodding to last year's strike by the California Nurses Association.
"A lot of us say that we're not going to be replaced by AI, we want to believe that, but in medicine, we can't just invoke this and expect that to address the underlying drivers as to why people are asking about this," Chang said on Tuesday's health care panel.
What could those underlying drivers be? I found some clues in Wolters Kluwer Health's new generative AI readiness report, released at the start of the month. Their survey of health care stakeholders found that 76 percent cite "reducing clinician burnout" as a major priority, and 85 percent say "recruiting/retaining nursing staff" is top of mind.
But only 45 percent of nurses responded "yes" when asked if generative AI can reduce clinician burnout.
I asked Dr. Peter Bonis, chief medical officer at Wolters Kluwer Health, to help me make sense of that gap—or "disconnect," as he called it.
"Our survey is indicating that there's an opportunity to work with allied health professionals and clinicians to deeply understand their needs and where some of these technologies can help; to have them on participatory boards as they themselves get educated on what these tools can do; and to select these tools so that they are optimizing their workflows, and they have agency in this process," he said.
I've been reporting on a lot of AI-related "disconnects" lately, both within health systems' AI deployment efforts and in external communications with vendors. I asked Bonis: Do these lapses indicate that we're taking the wrong approach to AI deployment in the health care industry?
He told me that this isn't exclusive to the health care industry, and that every business is wrestling with the same sort of issues. (Phew.) But, he acknowledged, health care is a high-stakes game, and it's important to deploy AI safely (and transparently) for the benefit of employees, patient care and overall health equity.
He believes health systems will succeed if they focus on patient care and bolster that foundation with sound operations and a successful business model.
"The fusion of those two directives is what creates a future-ready health care system that understands how to use these advanced technologies to advance their operations—to do that thoughtfully—and then to have a coherent pathway to start to use these tools to advance that higher stakes domain," Bonis said, "and that's the journey that we're on."
I also spoke with Dr. Perlin, head of the Joint Commission, about that journey to a "coherent" AI pathway. Read on to the Pulse Check section to see what he said.
Essential Reading
Aiming to improve the prior authorization process, HHS Secretary Robert F. Kennedy, Jr., and CMS Administrator Dr. Mehmet Oz hosted a roundtable of health insurance executives and stakeholders on Monday.
on Monday. Attendees agreed to six reforms: (1) standardizing electronic prior auth submissions, (2) reducing the volume of services that require prior auth, (3) honoring existing approvals during insurance transitions, (4) improving transparency and communication around decisions, (5) implementing real-time approvals for most requests by 2027 and (6) ensuring medical professionals review all denials.
(1) standardizing electronic prior auth submissions, (2) reducing the volume of services that require prior auth, (3) honoring existing approvals during insurance transitions, (4) improving transparency and communication around decisions, (5) implementing real-time approvals for most requests by 2027 and (6) ensuring medical professionals review all denials.
These companies were included in the discussion: Aetna, AHIP, Blue Cross Blue Shield Association, CareFirst BlueCross BlueShield, Centene Corporation, The Cigna Group, Elevance Health, GuideWell, Highmark Health, Humana, Kaiser Permanente and UnitedHealthcare. Together, they represent about 75 percent of Americans with commercial or Medicare Advantage plans.
The Lown Institute has released its highly anticipated index of America's Most Socially Responsible Hospitals. This year's honor roll comes at a critical time, as hospitals work to maintain equitable care amid potential Medicaid cuts, rising costs and ongoing workforce challenges. Duke Regional Hospital topped this year's acute care ranking, marking its fifth year on the list.
Eli Lilly's once-weekly insulin efsitora displayed promising results in Phase 3 clinical trials, reducing A1C and meeting safety standards for adults with Type 2 diabetes, according to detailed datapublished by the company this week. The new drug hopes to simplify diabetes management by reducing the frequency of insulin injections. Diabetes is becoming more prevalent in the United States , affecting nearly 15 percent of adults. By the end of 2025, Eli Lilly plans to submit the drug to global regulatory agencies for the treatment of Type 2 diabetes.
reducing A1C and meeting safety standards for adults with Type 2 diabetes, according to detailed datapublished by the company this week. The new drug hopes to simplify diabetes management by reducing the frequency of insulin injections.
Artisight, the NVIDIA-backed health tech company specializing in AI-powered "smart hospital" infrastructure, announced a $40 million investment from a dozen health systems. It's an unprecedented level of support from some of the nation's largest, most forefront integrated systems and academic medical centers. (The list of names was reviewed by Newsweek but is not being released to the public at this time.)
Pulse Check
Dr. Jonathan Perlin is president and CEO of the Joint Commission.
Dr. Jonathan Perlin is president and CEO of the Joint Commission.
Joint Commission
If you've been paying attention to the news lately, you may have had the same question that I did: What on earth is going on at the Joint Commission?The independent health care accreditation and certification organization has launched a couple high-profile, high-tech partnerships in recent weeks. First, it announced a long-term relationship with Palantir, intending to use the company's AI platform to streamline accreditation/certification processes. Then, it joined forces with the Coalition for Health AI to establish a "suite" of AI best practices playbooks and a new certification for hospitals. Now, if you work at a hospital, you likely live by the Joint Commission's standards. That's why I called Dr. Jonathan Perlin, president and CEO of The Joint Commission, last Friday. There was a bit of "geeking out about AI," as Perlin put it. But mostly, we discussed the recent CHAI partnership—and what it could mean for quality/safety standards and hospital certifications. Editor's Note: Responses are lightly edited for length and clarity.
AI is so different from other components used to assess quality. Each health system uses it in a unique way, and applications vary between hospitals, departments and even patient populations. "Good AI" can be tough to quantify. How do you plan to create a standard with this new certification program? CHAI's lane is really the technology itself, and ours is the organization's governance process for the responsible use of that. You may have seen our Responsible Use of Health Data Certification that has six attributes, and this is really an extension of that. What we anticipate—and this is a work in progress—is that building from the Responsible Use of Health Data [Certification], there'd be requirements for de-identification or privacy that could be data controls for security. There should be some mechanism for transparency with patients. Most importantly, there would be like an oversight or governance structure that addresses the algorithm's or the AI's performance. The notion is that an organization can look to CHAI and to the market to identify an AI tool, but it has to have an active and ongoing governance process to look at the performance of that tool in their environment. To give an example, I think there are three essential components. One is technical: Is [the tool] valid and reliable in a sort of mathematical sense? Second, is it valid and reliable clinically? Does it present the right clinical information? And third, is it valid and reliable in a demographic sense, that you're not applying an AI trained specifically for detection of sepsis in adults to children. To make that clear, if CHAI's lane is really the external performance of the algorithm and the assurance aspects outside of health care, the way we do this [new certification] is not specifically directed at the certification of the AI tool, but the certification of process for the organization's own governance and oversight of the use responsible use of that AI tool.
Will this certification assess tools that health systems developed internally, vendor tools that they deploy, or a combination of both? What else will you be looking at within each hospital's AI ecosystem? We anticipate that the certification, which would be given to health care organizations, would be based on the governance structure and the oversight structure I described [above]. We expect it would be applied both to homegrown and off-the-shelf technologies. Our focus is on continuous governance. Let me give an example that's literally closer to home. We just finished a renovation [at my house], and we had an electrical inspection after the work was completed. The wiring of the house is not going to change over time, but the wiring of AI, if it's retrained, if it drifts, etc., may change over time—so the organization needs to have a mechanism for periodic review of the performance of its "electrical system," to use the analogy, not just at inception, but periodically, or frankly, for the life of the use of that technology.
Any advice for health systems that are currently building up their AI governance structures, to ensure they're on the right path ahead of the Joint Commission and CHAI's certification?
Take a look at our Responsible Use of Health Data Certification , because it really provides insight into the concepts of governance as the regulatory frameworks are emerging. Despite the fact that device drug approvals are static, they are viewing the use of device (good outcomes or bad) as the responsibility of the clinicians and health care organizations that use those. Having come from operations and large systems myself, it's really important to have a set of externally validated standards that demonstrate what "good" looks like for responsible governance and oversight. I think organizations like ours are hugely excited about the potential, but we want to set up common standards to assure that we realize that potential responsibly.
If you liked this sneak peek, remember to check out next week's edition, which will include more of my interview with Perlin.
C-Suite Shuffles
Joseph Impicciche is retiring as CEO of Ascension, after six years at the helm of the St. Louis-based system. Eduardo Conrado, the health system's current president, will become its new CEO on January 1, 2026. Conrado was named to Ascension's executive team in 2018, after five years on its board of directors. Throughout his tenure, he has also served as the system's chief digital officer and chief strategy and innovation officer.
The Medical University of South Carolina (MUSC) has selected Dr. John Marymont as its next provost and executive vice president for academic affairs. Currently, he serves as vice president for medical affairs and dean of the medical college at the University of South Alabama.
Lovelace Health System is undergoing its fourth CEO change since 2022, the Albuquerque Journal reported. President and CEO Troy Greer resigned last week after two years in the role. The health system—one of the largest in New Mexico—declined to comment on his exit.
Executive Edge
Dr. Leigh Vinocur is a thought leader on stress management amongst health care professionals.
Dr. Leigh Vinocur is a thought leader on stress management amongst health care professionals.
Dr. Leigh Vinocur
Dr. Leigh Vinocur tells me that she considers herself a "lifelong learner." She's spent her career learning different elements of medicine, starting as a urology resident, becoming a board-certified emergency physician and serving as chief medical director for a major health system, overseeing more than 100 providers in the mid-Atlantic region. Now, she is the medical director of a men's health clinic and works part-time in clinical trials at a nutraceutical company. But throughout her working life, Vinocur has also learned a great deal about stress. After leaving her big-box health care role, she dealt with a "corporate medicine hangover." On July 12, she's releasing a book about her journey: Never Let Them See You Sweat: How Science Can Help Us Harness Stress for Success. This week, I connected with Vinocur to learn what her research—and personal experiences—taught her about stress in the health care setting. Here's what she told me: Editor's Note: Responses are lightly edited for length and clarity. "Today, in this political climate, we're seeing changes to health care, erosion of public health. It's making it even more difficult [to work in health care leadership]. There were always issues as a physician, fighting with insurance companies—but as physician executives, it's that kind of double bind that they're in, because they're caregivers, but they're administrators. They have rules they're enforcing, fiscal and institutional constraints, but they have to put their patients first, too. And it's stressful.
"Not all stress is always horrible and bad. It gets you to your tiptop performance, you know. Stress was an evolutionary development and advantage to keep us safe. If you were out there being chased by a predator, all those reactions from the hypothalamus, the pituitary, the adrenal release the cascade of hormones for that fight or flight. Whether you're a runner in the Olympics standing in the blocks, or whether you're an ER doctor waiting in the resuscitation room for those accident victims to come in, that little boost of stress gets you at your top performance. It's just this continued stress that is so challenging.
"I tell executive leaders that they need some buffer in between meetings. You need to create little micro-breaks during the day that are just for you to calm down. Whether that's meditation (there are apps on our watches and our phones), deep breathing (like box breathing, where you inhale through your nose for four seconds, hold it for four seconds and exhale for four seconds through your mouth), or leadership mentoring (where you have open dialogue and create a safe space to talk about some of the ethical dilemmas you may be facing).
Whether that's meditation (there are apps on our watches and our phones), deep breathing (like box breathing, where you inhale through your nose for four seconds, hold it for four seconds and exhale for four seconds through your mouth), or leadership mentoring (where you have open dialogue and create a safe space to talk about some of the ethical dilemmas you may be facing). "All throughout the book, there are discussions on things you can do in nature, like 'forest bathing.' Study after study says that if you can get to a green space, like a park, that can lower your blood pressure. Just being out in nature, getting outside—you don't even have to exercise—can be a great relief."
This is a preview of Access Health—Tap here to get this newsletter delivered straight to your inbox.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Forbes
2 hours ago
- Forbes
Healthcare's Denial Crisis Is Getting Worse—Unless Clinics Get Smarter With AI
Sally Ragab, Founder & CEO @ Neunetix. In early 2024, one of our clients, a specialty clinic in the Midwest, discovered it had unknowingly failed to submit over 6,000 claims due to a silent software integration failure. The result: nearly $200,000 in lost revenue, most of it unrecoverable due to expired filing windows. The culprit wasn't negligence or bad billing—it was fragmentation. It can create blind spots—leading to claim submission failures, missed deadlines, and preventable denials. Denied claims siphoned an estimated $260 billion from U.S. healthcare providers in 2024 alone. Specialty clinics—oncology centers, orthopedic practices, infusion providers—are hit hardest. Their margins are tight, their billing is complex and each denial cuts twice: once financially, and again in delayed or denied patient care. That's why the next wave of financial recovery in healthcare isn't coming from new payer contracts—it's coming from smarter tech. And AI, if used efficiently, is leading the way. The Real Cost Of Denials—And The Role Of AI Most providers still rely on a patchwork of systems to track, submit and appeal claims. When those systems don't integrate—or worse, when no one notices—they bleed revenue. AI can't fix everything. But it can predict denials before submission, help identify errors and even generate appeal language. When trained on the right data, it works fast, consistently and at scale. Our examination of 27 peer-reviewed and industry studies published between 2020 and 2025 found that models using advanced algorithms achieved AUROC scores between 0.82 and 0.92, which means it has an excellent predictive ability. Clinics that deployed those tools saw denial rates drop by up to 45%, shaving a median nine days off accounts receivable cycles and preserving roughly $1.4 million in annual savings per 100-bed equivalent each year. In short, today's off-the-shelf machine-learning tools already catch about nine out of 10 risky claims; exotic deep-learning models can do slightly better, but only if you feed them a million-plus past claims—far beyond what a typical clinic keeps on file. Why Specialty Clinics Feel The Pain First Specialty clinics face a perfect storm of challenges: high-cost drugs, evolving payer policies and thousands of diagnosis and procedure codes that need to match up precisely. One denial can equal a $10,000 loss. If one ICD-10 or HCPCS code is wrong—or out of date—the claim can be denied. And those codes change every year. When those denials stack up, clinics risk closing programs or closing entirely Adding to the chaos is the fact that insurers now use their own AI models to reject claims—models that providers often can't inspect. That creates an urgent need for transparency and speed on the provider side. AI won't eliminate denials entirely. But it can help level the playing field. A Five-Step Framework For AI Denial Management To help providers operationalize AI safely and effectively, I recommend this implementation roadmap: 1. Readiness Audit: Start by cataloging your denial categories, A/R metrics and critical claim data (CPT, HCPCS, ICD-10, prior-auth status). Dirty or missing 837/835 data accounts for up to 40% of predictive model errors. 2. Model Selection: Start simple. Boosted-tree models offer strong early performance and are interpretable—so your billing staff can see why a claim is flagged. Avoid 'black box' AI in healthcare. 3. Pilot And Back-Testing: Test the model in 'shadow mode' on a subset of claims for at least 90 days. Target a ≥20% precision lift before considering deployment. 4. Workflow Integration: Embed denial predictions directly into your EHR or billing software. Don't send billers hunting. Route high-risk claims to seasoned billers for pre-submission scrub or immediate appeal. 5. Continuous Governance: Retrain models quarterly. Audit for bias. Document your oversight. States like Connecticut are already drafting legislation that could require human review of AI-influenced decisions. Be ready. Ethics, Transparency And The Provider's Role Let me be clear: AI should not be used to deny care. Our mission at Neunetix is to protect access to care by helping providers recover revenue they've rightfully earned. That means appealing denials—not rubber-stamping them. This distinction matters. A 2025 AMA survey found that 60% of physicians worry AI could be used to systematically deny needed treatment. Transparency, override logs and clinician review should be standard practice. Done right, AI isn't a barrier to care—it's a lifeline for providers. The Bottom Line Specialty clinics don't have time to waste. With denial rates rising and payer rules getting harder to track, it's no longer optional to adopt smarter tools. It's essential. The clinics that embrace AI now—not as a magic fix, but as a smart assistant—will recover faster, protect more patient care and outperform peers still stuck in reactive mode. Knowing why you're getting denied isn't enough if you can't act on it quickly. And if your systems aren't talking to each other, neither are your dollars. Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
Newsweek
4 hours ago
- Newsweek
Map Shows States With Highest Use Of Harmful Pesticides
Based on facts, either observed and verified firsthand by the reporter, or reported and verified from knowledgeable sources. Newsweek AI is in beta. Translations may contain inaccuracies—please refer to the original content. Midwestern and Southern states use much higher rates of the potentially harmful pesticide 2,4-D, according to U.S. Geological Survey (USGS) data. A map by USGS, based on the latest data, collected in 2019, shows that use is most prevalent across Iowa, Missouri, Kansas, Illinois, Indiana, Ohio, Kentucky, with other states also using the herbicide in certain areas. 2,4-D was found to be "possibly carcinogenic to humans" in a 2015 report by International Agency for Research on Cancer (IARC) in the World Health Organization (WHO), although the report states there was "inadequate evidence" at the time to determine the extent of its harm. But there are justified concerns about potential harm, depending on exposure levels to 2,4-D. Newsweek has contacted the Department of Agriculture (USDA) via email for comment. Why It Matters 2,4-D has been used as a pesticide since the 1940s, and gained notoriety as one of the components of Agent Orange, the defoliant used by U.S. forces during the Vietnam War to destroy forestry and crops. While the other component of Agent Orange, 2,4,5-T, contained high levels of dioxin which was found to cause cancer and other health problems, dioxins are reportedly not found at detectable levels in 2,4-D products sold and used in the United States, USDA said. However, there is still some concern about the safety of its use, particularly as it has been used more widely in the last decade. A map to show the use of 2,4-D pesticides in different US states. A map to show the use of 2,4-D pesticides in different US states. Uncredited/U.S. Geological Survey What To Know 2,4-D is used to kill broadleaf weeds in various places, from lawns to fruit and vegetable crops, according to USDA. "The Midwest, Great Plains, and Northwestern U.S. have the highest 2,4-D usage, largely because these regions are the primary producers of corn, soybeans, wheat, and other field crops that are commonly treated with 2,4-D," Gurumurthy Ramachandran, director of the Johns Hopkins Education and Research Center for Occupational Safety and Health, told Newsweek. The increase in its use has been primarily due to the "rise of glyphosate-resistant weeds," Ramachandran said. Glyphosate, a once dominant herbicide, has become less effective as many weed species have developed resistance to it. To combat these resistant weeds, farmers have turned to alternative herbicides like 2,4-D, given its effective and low-cost control of broadleaf weeds. However, while USDA reported that "2,4-D generally has low toxicity for humans," others may disagree. The herbicide has been linked to significant health concerns, including non-Hodgkin's lymphoma (NHL), endocrine disruption, a decrease in fertility and increase in birth defects, according to the nonprofit organization Natural Resources Defense Council (NRDC). The IARC report detailed that there was "strong evidence that 2,4-D induces oxidative stress, a mechanism that can operate in humans, and moderate evidence that 2,4-D causes immunosuppression, based on in vivo and in vitro studies." "However, epidemiological studies did not find strong or consistent increases in risk of NHL or other cancers in relation to 2,4-D exposure," it added. The IARC also told Newsweek that the classification does not indicate the level of risk associated with a given level or circumstance of exposure. "The cancer risk associated with substances or agents assigned the same classification may be very different, depending on factors such as the type and extent of exposure and the degree of the effect of the agent at a given level of exposure," the organization said. The issue is that pesticides "can remain on or in food, and chronic dietary exposure has been linked to increased risks of metabolic syndrome, cancers, and other health problems," Ramachandran said. They can also "contaminate water, air, and soil, potentially affecting people living near treated fields or those exposed through drift and runoff," he added. This means that's its important for Americans to "wash all fruits and vegetables thoroughly before consumption; consider choosing organic produce to reduce dietary pesticide exposure, especially for children and pregnant women; and avoid entering fields or areas recently treated with pesticides and follow posted warnings," Ramachandran said. What People Are Saying Cynthia Curl, a professor in the School of Public and Population Health at Boise State University, Idaho, told Newsweek: "I think we need to be judicious in our use of these agricultural tools, understanding that exposures to these chemicals have been associated with adverse health effects. We should be supporting federal programs like the National Institute of Environmental Health Sciences, the United States Department of Agriculture, and the Occupational Safety and Health Administration that fund research, outreach and education to evaluate pesticide safety and support interventions to reduce pesticide exposures to workers, people living in agricultural communities and the general public." David F. Goldsmith, an occupational and environmental epidemiologist at Milken Institute School of Public Health, George Washington University, Washington, D.C, told Newsweek: "I am concerned if farmers or farmworkers are not using effective safety gear and thus may be excessively exposed via inhalation or skin contact. I am also concerned that 2,4-D may contaminate drinking water sources. Although, I believe that there is a direct risk for people who buy produce from fields that have had 2,4-D used on them." He added: "The public needs to be informed and vigilant about the use of herbicides, keeping them away from children, schools and using protective equipment such as gloves, face masks, and goggles. For farm workers, many of whom do not speak English, the information on the Material Safety Data Sheet needs to be translated and the farm managers who are sending them out into the fields need to ensure they have personal protective equipment, that they wear long sleeves and gloves." Gerald LeBlanc, a professor in the Department of Biological Sciences, North Carolina State University, told Newsweek: "I am not concerned that 2,4-D may cause cancer in humans. IARC has classified 2,4-D as a Group 2B carcinogen, which means that it is possibly carcinogenic to humans. In my estimation, 2,4-D might cause cancer in humans, but only at unrealistically high exposure levels." He added: "It's essential to recognize that toxicity is just one side of the coin when assessing chemical risk. The other side is exposure. Pesticide applicators are exposed to much higher levels than consumers of the agricultural products. Therefore, pesticide applicators may be at risk while the pesticide may pose no risk to consumers. The EPA sets limits on the amount of pesticides that can be found on agricultural products. As long as these limits are not exceeded, the pesticides are considered to be safe for human consumption. This is the case for 2,4-D, where residues associated with crops are typically well below levels that pose a risk of harm to consumers." What's Next Before the extent of the health impacts of 2,4-D are made clear, more research is needed and various governmental agencies are continuing to review the herbicide's safety. It's also important to note, the USGS map is based on 2019 data, the latest data collected by the agency, so the prevalence and use of the pesticide may have changed in more recent years. More recently, farmers have been finding "new replacements" for the pesticide, Diana S. Aga, a professor of chemistry at University at Buffalo, New York, and fellow of the American Chemical Society, told Newsweek. The USGS also told Newsweek that final annual pesticide-use estimates, for approximately 400 compounds, from 2018 to 2022 will be published later this year. "The pesticide use data releases are currently undergoing peer review, and we anticipate these products will be published around September 2025," they said.
Newsweek
4 hours ago
- Newsweek
Joint Commission CEO Breaks Down New AI Certification
Based on facts, either observed and verified firsthand by the reporter, or reported and verified from knowledgeable sources. Newsweek AI is in beta. Translations may contain inaccuracies—please refer to the original content. This is a preview of Access Health—Tap here to get this newsletter delivered straight to your inbox. I've settled on my word of the week: transparency. Over the last seven days, my schedule has looked like a patchwork quilt, spanning several distinct sectors of the health care industry. But from the government's roundtable on prior authorization reform to my conversation with Joint Commission CEO Dr. Jonathan Perlin (read on for more on both of those), transparency has emerged as a common thread. Health systems and insurance companies must be transparent with one another; health systems and insurance companies must be transparent with patients. And, of course, health care companies must be transparent with their employees and customers—especially when deploying AI. Monday through Wednesday, many of my colleagues were in sunny Sonoma, California, for Newsweek's cross-industry AI Impact Summit. Senior Reporters Lauren Giella and Katherine Fung kept eyes on the health care side of the agenda, which included speakers from Kaiser Permanente, Hospital for Special Surgery, AdventHealth and UMass Medical. "Transparency was a big theme for AI adoption in health care—not only when dealing with medical records and personal data, but also for why and how organizations are implementing automation tools," Giella wrote. Dr. Allen Chang, ACMIO at UMass Medical, warned the audience not to neglect employees' concerns about losing jobs to AI, nodding to last year's strike by the California Nurses Association. "A lot of us say that we're not going to be replaced by AI, we want to believe that, but in medicine, we can't just invoke this and expect that to address the underlying drivers as to why people are asking about this," Chang said on Tuesday's health care panel. What could those underlying drivers be? I found some clues in Wolters Kluwer Health's new generative AI readiness report, released at the start of the month. Their survey of health care stakeholders found that 76 percent cite "reducing clinician burnout" as a major priority, and 85 percent say "recruiting/retaining nursing staff" is top of mind. But only 45 percent of nurses responded "yes" when asked if generative AI can reduce clinician burnout. I asked Dr. Peter Bonis, chief medical officer at Wolters Kluwer Health, to help me make sense of that gap—or "disconnect," as he called it. "Our survey is indicating that there's an opportunity to work with allied health professionals and clinicians to deeply understand their needs and where some of these technologies can help; to have them on participatory boards as they themselves get educated on what these tools can do; and to select these tools so that they are optimizing their workflows, and they have agency in this process," he said. I've been reporting on a lot of AI-related "disconnects" lately, both within health systems' AI deployment efforts and in external communications with vendors. I asked Bonis: Do these lapses indicate that we're taking the wrong approach to AI deployment in the health care industry? He told me that this isn't exclusive to the health care industry, and that every business is wrestling with the same sort of issues. (Phew.) But, he acknowledged, health care is a high-stakes game, and it's important to deploy AI safely (and transparently) for the benefit of employees, patient care and overall health equity. He believes health systems will succeed if they focus on patient care and bolster that foundation with sound operations and a successful business model. "The fusion of those two directives is what creates a future-ready health care system that understands how to use these advanced technologies to advance their operations—to do that thoughtfully—and then to have a coherent pathway to start to use these tools to advance that higher stakes domain," Bonis said, "and that's the journey that we're on." I also spoke with Dr. Perlin, head of the Joint Commission, about that journey to a "coherent" AI pathway. Read on to the Pulse Check section to see what he said. Essential Reading Aiming to improve the prior authorization process, HHS Secretary Robert F. Kennedy, Jr., and CMS Administrator Dr. Mehmet Oz hosted a roundtable of health insurance executives and stakeholders on Monday. on Monday. Attendees agreed to six reforms: (1) standardizing electronic prior auth submissions, (2) reducing the volume of services that require prior auth, (3) honoring existing approvals during insurance transitions, (4) improving transparency and communication around decisions, (5) implementing real-time approvals for most requests by 2027 and (6) ensuring medical professionals review all denials. (1) standardizing electronic prior auth submissions, (2) reducing the volume of services that require prior auth, (3) honoring existing approvals during insurance transitions, (4) improving transparency and communication around decisions, (5) implementing real-time approvals for most requests by 2027 and (6) ensuring medical professionals review all denials. These companies were included in the discussion: Aetna, AHIP, Blue Cross Blue Shield Association, CareFirst BlueCross BlueShield, Centene Corporation, The Cigna Group, Elevance Health, GuideWell, Highmark Health, Humana, Kaiser Permanente and UnitedHealthcare. Together, they represent about 75 percent of Americans with commercial or Medicare Advantage plans. The Lown Institute has released its highly anticipated index of America's Most Socially Responsible Hospitals. This year's honor roll comes at a critical time, as hospitals work to maintain equitable care amid potential Medicaid cuts, rising costs and ongoing workforce challenges. Duke Regional Hospital topped this year's acute care ranking, marking its fifth year on the list. Eli Lilly's once-weekly insulin efsitora displayed promising results in Phase 3 clinical trials, reducing A1C and meeting safety standards for adults with Type 2 diabetes, according to detailed datapublished by the company this week. The new drug hopes to simplify diabetes management by reducing the frequency of insulin injections. Diabetes is becoming more prevalent in the United States , affecting nearly 15 percent of adults. By the end of 2025, Eli Lilly plans to submit the drug to global regulatory agencies for the treatment of Type 2 diabetes. reducing A1C and meeting safety standards for adults with Type 2 diabetes, according to detailed datapublished by the company this week. The new drug hopes to simplify diabetes management by reducing the frequency of insulin injections. Artisight, the NVIDIA-backed health tech company specializing in AI-powered "smart hospital" infrastructure, announced a $40 million investment from a dozen health systems. It's an unprecedented level of support from some of the nation's largest, most forefront integrated systems and academic medical centers. (The list of names was reviewed by Newsweek but is not being released to the public at this time.) Pulse Check Dr. Jonathan Perlin is president and CEO of the Joint Commission. Dr. Jonathan Perlin is president and CEO of the Joint Commission. Joint Commission If you've been paying attention to the news lately, you may have had the same question that I did: What on earth is going on at the Joint Commission?The independent health care accreditation and certification organization has launched a couple high-profile, high-tech partnerships in recent weeks. First, it announced a long-term relationship with Palantir, intending to use the company's AI platform to streamline accreditation/certification processes. Then, it joined forces with the Coalition for Health AI to establish a "suite" of AI best practices playbooks and a new certification for hospitals. Now, if you work at a hospital, you likely live by the Joint Commission's standards. That's why I called Dr. Jonathan Perlin, president and CEO of The Joint Commission, last Friday. There was a bit of "geeking out about AI," as Perlin put it. But mostly, we discussed the recent CHAI partnership—and what it could mean for quality/safety standards and hospital certifications. Editor's Note: Responses are lightly edited for length and clarity. AI is so different from other components used to assess quality. Each health system uses it in a unique way, and applications vary between hospitals, departments and even patient populations. "Good AI" can be tough to quantify. How do you plan to create a standard with this new certification program? CHAI's lane is really the technology itself, and ours is the organization's governance process for the responsible use of that. You may have seen our Responsible Use of Health Data Certification that has six attributes, and this is really an extension of that. What we anticipate—and this is a work in progress—is that building from the Responsible Use of Health Data [Certification], there'd be requirements for de-identification or privacy that could be data controls for security. There should be some mechanism for transparency with patients. Most importantly, there would be like an oversight or governance structure that addresses the algorithm's or the AI's performance. The notion is that an organization can look to CHAI and to the market to identify an AI tool, but it has to have an active and ongoing governance process to look at the performance of that tool in their environment. To give an example, I think there are three essential components. One is technical: Is [the tool] valid and reliable in a sort of mathematical sense? Second, is it valid and reliable clinically? Does it present the right clinical information? And third, is it valid and reliable in a demographic sense, that you're not applying an AI trained specifically for detection of sepsis in adults to children. To make that clear, if CHAI's lane is really the external performance of the algorithm and the assurance aspects outside of health care, the way we do this [new certification] is not specifically directed at the certification of the AI tool, but the certification of process for the organization's own governance and oversight of the use responsible use of that AI tool. Will this certification assess tools that health systems developed internally, vendor tools that they deploy, or a combination of both? What else will you be looking at within each hospital's AI ecosystem? We anticipate that the certification, which would be given to health care organizations, would be based on the governance structure and the oversight structure I described [above]. We expect it would be applied both to homegrown and off-the-shelf technologies. Our focus is on continuous governance. Let me give an example that's literally closer to home. We just finished a renovation [at my house], and we had an electrical inspection after the work was completed. The wiring of the house is not going to change over time, but the wiring of AI, if it's retrained, if it drifts, etc., may change over time—so the organization needs to have a mechanism for periodic review of the performance of its "electrical system," to use the analogy, not just at inception, but periodically, or frankly, for the life of the use of that technology. Any advice for health systems that are currently building up their AI governance structures, to ensure they're on the right path ahead of the Joint Commission and CHAI's certification? Take a look at our Responsible Use of Health Data Certification , because it really provides insight into the concepts of governance as the regulatory frameworks are emerging. Despite the fact that device drug approvals are static, they are viewing the use of device (good outcomes or bad) as the responsibility of the clinicians and health care organizations that use those. Having come from operations and large systems myself, it's really important to have a set of externally validated standards that demonstrate what "good" looks like for responsible governance and oversight. I think organizations like ours are hugely excited about the potential, but we want to set up common standards to assure that we realize that potential responsibly. If you liked this sneak peek, remember to check out next week's edition, which will include more of my interview with Perlin. C-Suite Shuffles Joseph Impicciche is retiring as CEO of Ascension, after six years at the helm of the St. Louis-based system. Eduardo Conrado, the health system's current president, will become its new CEO on January 1, 2026. Conrado was named to Ascension's executive team in 2018, after five years on its board of directors. Throughout his tenure, he has also served as the system's chief digital officer and chief strategy and innovation officer. The Medical University of South Carolina (MUSC) has selected Dr. John Marymont as its next provost and executive vice president for academic affairs. Currently, he serves as vice president for medical affairs and dean of the medical college at the University of South Alabama. Lovelace Health System is undergoing its fourth CEO change since 2022, the Albuquerque Journal reported. President and CEO Troy Greer resigned last week after two years in the role. The health system—one of the largest in New Mexico—declined to comment on his exit. Executive Edge Dr. Leigh Vinocur is a thought leader on stress management amongst health care professionals. Dr. Leigh Vinocur is a thought leader on stress management amongst health care professionals. Dr. Leigh Vinocur Dr. Leigh Vinocur tells me that she considers herself a "lifelong learner." She's spent her career learning different elements of medicine, starting as a urology resident, becoming a board-certified emergency physician and serving as chief medical director for a major health system, overseeing more than 100 providers in the mid-Atlantic region. Now, she is the medical director of a men's health clinic and works part-time in clinical trials at a nutraceutical company. But throughout her working life, Vinocur has also learned a great deal about stress. After leaving her big-box health care role, she dealt with a "corporate medicine hangover." On July 12, she's releasing a book about her journey: Never Let Them See You Sweat: How Science Can Help Us Harness Stress for Success. This week, I connected with Vinocur to learn what her research—and personal experiences—taught her about stress in the health care setting. Here's what she told me: Editor's Note: Responses are lightly edited for length and clarity. "Today, in this political climate, we're seeing changes to health care, erosion of public health. It's making it even more difficult [to work in health care leadership]. There were always issues as a physician, fighting with insurance companies—but as physician executives, it's that kind of double bind that they're in, because they're caregivers, but they're administrators. They have rules they're enforcing, fiscal and institutional constraints, but they have to put their patients first, too. And it's stressful. "Not all stress is always horrible and bad. It gets you to your tiptop performance, you know. Stress was an evolutionary development and advantage to keep us safe. If you were out there being chased by a predator, all those reactions from the hypothalamus, the pituitary, the adrenal release the cascade of hormones for that fight or flight. Whether you're a runner in the Olympics standing in the blocks, or whether you're an ER doctor waiting in the resuscitation room for those accident victims to come in, that little boost of stress gets you at your top performance. It's just this continued stress that is so challenging. "I tell executive leaders that they need some buffer in between meetings. You need to create little micro-breaks during the day that are just for you to calm down. Whether that's meditation (there are apps on our watches and our phones), deep breathing (like box breathing, where you inhale through your nose for four seconds, hold it for four seconds and exhale for four seconds through your mouth), or leadership mentoring (where you have open dialogue and create a safe space to talk about some of the ethical dilemmas you may be facing). Whether that's meditation (there are apps on our watches and our phones), deep breathing (like box breathing, where you inhale through your nose for four seconds, hold it for four seconds and exhale for four seconds through your mouth), or leadership mentoring (where you have open dialogue and create a safe space to talk about some of the ethical dilemmas you may be facing). "All throughout the book, there are discussions on things you can do in nature, like 'forest bathing.' Study after study says that if you can get to a green space, like a park, that can lower your blood pressure. Just being out in nature, getting outside—you don't even have to exercise—can be a great relief." This is a preview of Access Health—Tap here to get this newsletter delivered straight to your inbox.
