The NHCAA 2025 Annual Training Conference, held November 18-21, featured discussions on the use of AI in investigations, how bad actors exploit the technology, common durable medical equipment fraud schemes, and Affordable Care Act enrollment fraud.
AI has been a major trend in AI this year, said NHCAA CEO Lou Saccoccio, and it featured heavily in ATC sessions in terms of its impact on health care fraud investigations and how bad actors are using the technology.
Saccoccio shared other highlights from the conference, as well as the NHCAA’s goals for 2026.
How did the annual training conference go? Could you share some highlights?
The conference went very well. We had a bit of a challenge in that the government shutdown prevented our federal attendees from attending. But other than that, everything went very well and we were able to have them present via Zoom, so most of the sessions were covered and with presenters from the federal government.
One of our highlights was keynote speaker Kim Brandt from CMS, who gave us a good overview of what’s happening at CMS right now (and) some of their priorities, the biggest being “rooting out fraud,” which is a phrase they’re using now. Dr. Oz is a CMS administrator; it is one of his priorities. Tthe hat was a highlight.
The general session we had brought together a panel of federal and state figures who play a vital role in the fight against fraud. We brought in the HHS Deputy Inspector General for Investigations. We heard from the head of the FBI’s health care fraud unit. We had someone from the Department of Justice and someone from a state Medicaid fraud control unit. It went very well. They talk about their successes, their priorities for the future and some of the challenges they face.
And then another highlight, I think, was a general session that we had on artificial intelligence, which is a major issue for everyone, but it certainly relates to the area of fraud – whether AI can be used to commit fraud, which is certainly the case with all kinds of ways of imitating individuals’ voices and medical records and even X-rays and those sorts of things – but also the opportunities for using AI to detect and prevent fraud. So this is a theme that was also discussed throughout the conference: the impact that artificial intelligence is going to have on healthcare fraud.
Many sessions focused on AI and fraud. How would you say investigators use AI to detect and prevent fraud?
Well, we’re still at the beginning, very early in the process as to how they can use it to prevent fraud. Obviously you have to apply AI to the claims data and try, again, to find the outliers, but even doing that you’re still going to have to do an investigation, so there’s no silver bullet. But it can certainly help prioritize and perhaps prevent some fraud in the sense that you could stop payment for certain things before payment is made, or subject claims to a more rigorous review process that appear to be outliers, which can help prevent fraud.
And then there are some administrative things. Let’s say phone calls come in alleging that fraud has been committed. AI can play a role in handling these calls and other leads you may have. I think it’s still very early in the process as to how this can be used to assist, detect and prevent fraud.
You mentioned that bad actors use AI to manipulate their voices. Are there other tactics they use to try to avoid detection? And do these tactics complicate investigators’ investigations?
They are starting to complicate investigations. As I mentioned, medical records are a major problem – they generate fake medical records, cut and paste, that sort of thing, so that if they turn them over, they can look like real medical records.
People looking to commit fraud could also use AI to try to uncover vulnerabilities in payer rules to determine what might work and what might not. I think even for them, this is a new process as well, so it will be up to whoever is ahead of the curve in terms of artificial intelligence as it relates to healthcare.
Other sessions We were looking at fraud around Affordable Care Act enrollment, which I thought was a particularly relevant session given that enrollment is now open. Can you talk about it?
One of the big problems with ACA enrollment is brokers and people who enroll people from different states to try to make money to enroll those people. So this is a known problem, and AI could help by identifying registered people.
There are also frauds involving people pretending to have limited income to qualify for grants when they may not really be eligible. It’s more of a government issue, because it’s the government that pays the subsidies.
So anytime you have large data sets like that, AI could play a role in trying to discern what’s real and what’s not, or at least give investigators a head start on where to look and what further investigations to conduct.
There has been recent talk in Washington of subsidy fraud. Do you think there is anything at the policy level that could help prevent registration fraud? Is there anything the federal government can do?
I know there are discussions in Congress about possibly expanding the enhanced subsidies that were in effect during COVID, but in doing so also limiting the number of people who qualify based on their income. So if your income is above a certain level, should there really be subsidies, or should there really be enhanced subsidies? So I think that comes into play in any discussions that might take place in Congress regarding extending these grants.
And then, to prevent fraud, I guess we need to do a better job of verifying eligibility. Whether it’s Medicaid or subsidies, there are rules, and are those rules enforced? And what efforts are you putting into trying to actually verify the eligibility of these people? Do you just take what they tell you at face value, or do you do some checking? Obviously, this could be a very intense and expensive process in itself, so you need to find the balance between thorough verification and making the program work for the people it’s supposed to work for.
Mr.More generally, what have been the biggest trends in healthcare fraud this year?
First of all, of course it will be AI. We formed an AI working group, and based on that working group, we wrote a brief on the use of generative AI in healthcare fraud. So that was a big part for us. When it comes to general healthcare fraud, durable medical equipment remains a problem. This has always been a problem and it will continue to be a problem.
Earlier this year and late last year, there was the issue with catheters: people were contacted by phone and provided information, and then doctors signed an EMR that wasn’t needed, and they never saw the patients. Telemarketing fraud is therefore significant.
Another issue that I think comes into play is behavioral health care fraud. A lot of (behavioral health care) is delivered virtually, which is good and important, but it lends itself to fraud. …Remote patient monitoring is another issue coming to the forefront. … I read something today where the former CDC director – she was only in that position for a short time – but she estimates that as we move down this path, about 90% of health care will be delivered in homes rather than in hospitals.
Again, because you’re using technology, there’s always a risk of fraud. Every time there is an advancement, of some nature, there will be people who find a way to use it to discover the vulnerabilities of that advancement in order to commit fraud. So those are some of the things, but in all kinds of areas: lab testing – you could go down the list – opioid problems with pill mills, doctors writing prescriptions basically for money, that sort of thing. Although I think the federal government has done a really good job, both at the federal level and at the state level, of really making a dent in this problem.
Looking forward to next year, what will be the NHCAA’s priority?
For us, we really need to focus on the education and training side. A big deal for us will be a new educational platform that we implemented just two years ago, but is starting to grow: NHCAA Learn. We’re trying to make sure that we increase the number of on-demand courses that we offer on this learning management system and open it up, as we have, to our members, essentially as a member benefit. (We’re) also trying to get subscriptions with the government so that on an annual basis they pay a certain amount and everyone – let’s say the FBI, the HHS OIG – has access to it.
We currently have a number of courses available on the site, but one of our main goals over the next year is to expand the library of courses available and make it a key tool for our members, both our commercial members and our government partners, to have a learning platform that is useful to them and provides them with what they need to train their employees.