KPMG’s Ash Shehata discusses AI excitement and uncertainty | ViVE 2025

Nashville - Ash Shehata sees the healthcare industry from a variety of angles.

Shehata, KPMG’s U.S. sector leader for health care, moderated a discussion at the opening of the ViVE digital health conference on the claims management process, its problems, and the prospects of artificial intelligence in improving the process.

After the session, he sat down with Chief Healthcare Executive® to talk about AI, the undeniable enthusiasm around its use in healthcare and lingering questions. He also discussed digital health investments, the divide between payers and providers, academic medical centers, and more. Here are some of the highlights.

Q: What type of conversations are you having? What are you looking to talk about?

A: “You know, Gen AI is a really big topic right now. It was interesting, just from the moment you and I stepped out of the session, people were asking a lot of questions around where the targeted impact of Gen AI could be from an innovation cycle, because I think everybody sees the promise, but where are you going to get the biggest impact? I think that seems to be the big question right now.”

“And it was interesting in our session today, the one area that really did come out was about putting the content in the appropriate fields to submit a clean claim. And I know it sounds like something we should have figured out 20 years ago. But the reality of it is, imagine if we could just use the technology that way, and there's a lot of underlying reasons, like claim systems and EHR systems are built fundamentally differently, but sometimes that information for a clean claim requires the longitudinal view of the patient or the member. And I think if we're smart about it, we can go back and pull EHR data that might be lab, radiology, ancillary services, to help justify the clinical case. So I think it's getting into those kinds of details that I'm hopeful in the next few days, some of the solutions we see will start to surface.”

(See part of our conversation in this video. The story continues below.)

Q: In that session, it was mentioned that AI is not a panacea for some of these problems, too. And I think we’re starting to see more and more that you have to fix some processes.

A: “I love the points of view. They were: engage physicians, fix processes, and don't just use AI and have this burden of complexity, because ultimately you'll just end up with an AI-enabled complex process, which isn't really a good idea. And people always ask me that question, like, do you simplify first before doing AI? Or the doctor says, sometimes I just need to fix the solution, so maybe sometimes I just go right at it. But at the end of the day, these enterprise solutions do need a major overhaul, I think, to be sustainable.”

Q: In health care, there's a growing appetite to use AI in a bunch of different ways. Do investors see that? Are they as bullish on it as the health industry, or are investors wanting to wait and see?

A: “It’s interesting. I think investors coming out of the JP Morgan conference, were still very much hell-bent on AI and tech is still going to get them the highest multiples of everything they've seen, just because of the promise of the technology. However, the money … the spigot hasn't started yet.

“So I would just say the anticipation is there, the future valuations are there, but the funds haven't started flowing yet, and I think there is some reticence still of whether it's an environmental question. There's certainly a lot of questions about the new administration. So a lot of things like that, I think, are still holding people back. But I think the promise of the technology is still keeping people pretty positive.”

Q: Much of the use of AI in healthcare is in the back end, or business operations. Are you starting to see more folks using AI in clinical areas, or are they still tentative?

A: “I’ve just done some recent board reports for some of our clients, and I would say the clinical AI use cases have been quite dramatic in the last six months. And there's a greater comfort, you know, people have understood the risk areas, understood maybe creating new governance structures, especially in academic medical centers. I think clearly they're the ones saying, this is something we do, and we're going to get ahead of it. We’ve got to get to it.

“I do think the administrative capabilities, however, are much more mature. So what we heard today about these enterprise solutions, if you think about if you're a startup company and you want to make an investment in new AI capability, that's likely going to get bought out by a big player, it will likely end up being administrative first and clinical second, because you can imagine there's a lot of other risks that have to be reviewed before we hit commercialization.”

Q: With academic medical centers, it seems like there's a lot of excitement about using AI in terms of research, in terms of mining through all the data they have, including decades of research in cancer or other diseases. Are you seeing a lot of enthusiasm among academic medical centers in using AI to help harness that research?

A: “Absolutely. You know, academic medical centers are really at the juncture of three areas. They're in this biopharma revolution, which is using precision health and precision medicine in new ways. They have the resources. They have the students coming out of the schools, not only the medical schools, but many of them have very good business schools that understand how to make investments in these areas. And they also have the technologists coming out of their schools. So they have that area. 

“And then the third area is they have the protocols. When we talk about the appropriate use of AI, whether it's around privacy or security or ethics, they have the educators who really study these areas, and have understood, you know, how to look at the risks appropriately. So I'm really, really bullish about, you know, putting academic medical centers in the middle of the future development of AI and healthcare.”

Q: Hospitals have talked about their frustrations with the rise in denials from insurers. Some of their frustration, from their perspective, has been with the use of AI tools leading to more denials. Over the last couple of years, we’ve heard that AI is supposed to streamline the claims everybody agrees should get paid. So how much of a problem is that from your perspective?

A: “I think one of the questions that came up (in the session) was, have we made a difference? And I think that some of the data was like, we haven't really made any difference. And I think the problem is it's becoming a zero sum game. As a health system might improve their ability to submit claims, the payers are also improving their ability to extract the information and deny claims. So we're not really gaining ground here together.

“So I think the biggest promise I heard today was, that there's a hope to work together between payers and providers, and even sometimes I'll call it the ‘pay-viders.’ So if you got a health system with your own payer, working better together as one system to create less stress, less frustration and more value, I think has got to be the way forward. We're at a zero sum game right now and we're spending more and more and getting probably less.”