The Department of Veterans Affairs (VA) has been working hard to reduce the amount of fraud the agency faces when it comes to veterans’ benefits claims, and better leveraging data and AI has been helping in that fight, said Erick Zenteno, program manager of Fraud Prevention Services at VA.
During GDIT’s “How AI can Prevent and Detect Fraud, Waste and Abuse” webinar, Zenteno described challenges VA has faced in preventing fraud, as well as areas of improvement.
“The biggest challenge for us currently is the integration [and] centralization of data,” Zenteno said. “We’re working on models, we’re working on approaches, solutions to better serve populations to prevent them from fraud. It’s only in the best interest of VA to get not only data, but real-time data that will only help the model get smarter and also be able to catch the fraud before it happens.”
Zenteno said that although AI can help the VA work towards getting real-time data, the agency is not fully there yet. However, he said they have implemented solutions regarding data access.
“I can categorize them [solutions] into four different buckets – first one being commercial best practices, second one being the technology, data resourcing, and the last one being a holistic approach,” Zenteno said.
“We have a framework to streamline the process to enable faster identification and resolution of fraud. We have shown significant reduction. But we can always do better,” he added.
One way the VA has reduced fraudulent payments has been by identifying all deceased veterans. By stopping payments to deceased veterans, Zenteno said his agency was able to enhance its data accuracy significantly.
Although identifying deceased veterans may seem like a straightforward solution, Zenteno said a lot of fraudulent activity isn’t so easy to detect. He said activities on the dark web are an increasing threat vector that bears more investigation.
“It is scary to know that things that could have been there that you are just currently not tracking for different reasons, but I do see the dark web being an area that where we need to start having one: discussions, two: efforts, and three: resources allocated … to be able to see what’s there,” he said. “Right now it’s this hidden black area, that I don’t know what’s under the hood.”