The Department of Transportation will use AI and advanced analytics to hunt down fraudsters and even carriers that employ illegal alien drivers, the department's Deputy Secretary Steven Bradbury announced on Monday.
As the keynote speaker at the Transportation Research Board's annual meeting, Bradbury gave great detail into the department's work to undo the previous administration's focus on diversity and climate goals, as well as dropping some hints about the future of trucking regulation.
"This administration is embracing AI, leaning into it," said Bradbury. "We're planning to use AI to accelerate our rulemaking process and really achieve much quicker turnaround in terms of the drafting of proposed rules, the digesting of comments, etc."
Not only that, the Federal Motor Carrier Safety Administration's longstanding manpower shortage in fighting fraud could get a lift from new tech, he said.
"We also want to use advanced data analytics to improve the accuracy and the effectiveness of our enforcement efforts," Bradbury added. "One example of that is to identify fraud in the trucking industry and identify instances where trucking companies are using illegal foreign drivers."
Bradbury echoed what thousands of Overdrive readers have reported in extensive surveying: Foreign drivers are bringing down rates.
[Related: Ban non-domiciled CDLs for foreign drivers? Owner-ops weigh in]
The use of "illegal foreign drivers" is "happening very extensively, unfortunately in the country and it's eating the lunch of American truckers because it's driving down to unreasonable levels the compensation for truckers," said Bradbury.
Bradbury said DOT also sought to reform the rulemaking process in support of "unleashing the power of American innovation."
But how might AI help stop fraud? After all, there's already an entire, highly technical "carrier vetting" industry that subjects carriers to all kinds of data questions to avoid fraud and cargo theft.
Additionally, FMCSA has already overhauled the registration system in an attempt to improve compliance and stem the tide of fake carriers using fake addresses.
Dale Prax, FMCSA's so-called "worst critic" and the head of the Freight Validate, a broker and carrier facing vetting tool, frequently talks to the agency's top brass and had some ideas about where AI could improve outcomes, including with:
- Imminent hazard out-of-service orders
- Principal Place of Business abuse
- Recycled phone numbers and undisclosed affiliations
- Identifying CDL mills
- Issuing carrier safety ratings
- Finding "illegal" drivers
Imminent hazard out-of-service orders, sparingly used by FMCSA, could get applied more quickly, Prax said, if FMCSA used AI to "identify the highest-risk carriers immediately using real signals: crash trends, OOS spikes, serious hours of service violations, equipment violations and inspection volatility" and then "place them out-of-service before the next crash."
Prax has long pushed FMCSA to crack down on carriers using P.O. boxes or registered agent locations as Principal Place of Business addresses, and made considerable progress on that front, but with AI, FMCSA "could do more," he said.
[Related: 'WTFFMCSA': Carriers mocking the agency through its own registration system?]
"AI could quickly identify recycled phone numbers, undisclosed affiliations with other FMCSA-regulated entities, discrepancies in identity between Secretary of State business filings and FMCSA data," and more, he said.
With regard to "chameleon carriers," Prax said "AI would connect hidden affiliations in minutes, not after a crash, a claim, or a victim."
Other blossoming initiatives like identifying CDL mills could benefit from AI too, he said.
"When a training/testing pipeline produces drivers who immediately rack up violations and OOS orders, that can be identified quicker with AI," said Prax.
Perhaps most applicable to AI would be issuing carrier safety ratings. Prax said he's talked directly with FMCSA about this use case.
"In December I spoke to FMCSA leadership specifically about how AI can speed up issuance of safety ratings," he said. "First, it could fast-track proven safe carriers (20 years in business with not so much as a speeding ticket) while also using roadside inspection patterns to flag high-risk carriers for immediate review."
With 94% of carriers operating lacking any safety rating at all, "this is vital right now," he said.
As the Supreme Court mulls a major case on broker liability regarding carrier selection, "we need AI to help give brokers data to look at rather than hold them accountable for something they have no concrete safety information to base their decisions on," said Prax.
When it comes to finding "illegal" drivers, Prax said there's a use case for that, too.
"I could see the agency using AI to compare immigration records and/or I-9 documents to CDLs as part of the audit process to determine if driver qualification records include undocumented workers," said Prax.
Overdrive asked DOT for more specific information about how it might use advanced tech to find illegal drivers or fraudsters, and will update this article or publish again if we hear back.
