Transportation Department sees modernization as key to fighting fraud

Modernization is a top priority for the Department of Transportation, whether it’s the multibillion dollar overhaul of the Federal Aviation Administration’s air traffic control system or part of the agency’s ongoing fight against fraud. 

In panel discussions Friday at ACT-IAC’s Imagine Nation ELC 2025 in National Harbor, Md., DOT officials spoke about how they’re leveraging technology to make agency operations more efficient and better deal with bad actors who have a clear advantage over federal tech professionals.

“One thing that is common,” said Ankur Saini, DOT’s acting chief product and technology officer, “is that the fraudsters have more information than you do.”

Saini’s teams are working to mitigate this issue by reaching a level of technical maturity and modernization that enables seamless data sharing and aggregation. 

It’s not an easy feat, Saini said, pointing to a lack of interoperability between data systems, capability disparity and the reengineering process. 

“We’re taking on a lot of technology initiatives and seeing how we can actually go in and equip people with the knowledge and ability to make those decisions … and close that gap,” Saini said. “Experimentation is key, and you have to have a learning mindset.”

Saini leans on change management techniques to ensure the approach is adopted throughout the organization, such as clear communication and identifying individuals who can help leadership facilitate the process. 

“Digital transformation doesn’t stop when you implement a system,” Saini said. “As a leader in the organization, you have to go out and say, ‘We are going to do things differently.’” 

In the next five years, Saini is aiming to consolidate a 60-plus application environment to just seven platforms and a data lake to help the agency better define and identify unsafe motor vehicles. AI will also likely play a role in the coming years, Saini said, by helping teams comb through troves of data and identify patterns. 

The stakes are high to stop fraud where the agency can. 

“Our goal is to make sure folks get home safely,” Saini said. “We see fraud as folks that should not be on the road … and the result is far more devastating than losing money. There’s an empty chair at the dinner table.”

A new era of AI-driven efficiency

Sometimes new technologies are a bit blinding for organizations, especially as agencies are asked to do more with less due to DOGE-led funding cuts and workforce reductions. 

Robotic process automation, or RPA, for example, was introduced to workers in federal agencies as a panacea, according to Neil Chaudhry, senior AI advisor at DOT. The hype surrounding the tool caused frustration as workers applied it to the wrong problems. 

“Now, we have a credibility problem as we try to bring in artificial intelligence solutions to help with government efficiency,” Chaudhry said during a separate panel discussion Friday. Caution and clarity are critical as federal technology leaders move forward on AI plans and aim to avoid friction. 

Federal agencies have ramped up AI efforts following the release of the Trump administration’s AI Action Plan, which pushed agencies to adopt emerging tech and partner with private companies. 

The traditional big tech mainstays and up-and-coming AI companies have been quick to jump on board and offer a helping hand, easing access by lowering costs. The Department of Health and Human Services took advantage of OpenAI’s $1 ChatGPT Enterprise offering in September. The Office of Personnel Management joined in, too, rolling out Microsoft Copilot and ChatGPT the same month. 

In its hunt for efficiency, DOT is transitioning away from legacy providers and rolling out Google Workspace with Gemini. The agency has about 12,000 employees onboarded and plans to add 40,000 more in the next year, according to a Friday announcement. 

The AI and collaboration software suite switch comes just months after the General Services Administration approved a OneGov deal that put Gemini for Government within agency reach for mere pennies. 

AI is part of the efficiency puzzle, Chaudhry said, but it isn’t the answer to every problem in federal IT. 

“Ultimately, efficiency [comes from] taking a step back, identifying the problem and then narrowing down the appropriate toolset to solve the problem,” Chaudhry said. “Change executed poorly is always bad.”

As agencies look for ways to improve efficiency, leaders who leave workers behind will likely find their efforts wasted, according to Chaudhry. 

“Until you show people how to use the tool in a way that meets their needs and supports their process, they’ll never use the tool,” Chaudhry said. “Or they’ll use it so inefficiently … any savings you had, you just lost.”