Freight industry leaders are warning that the rush to adopt artificial intelligence in logistics is outpacing the operational changes needed to make it work, with one expert saying only a small fraction of companies have redesigned their workflows around the technology.
The discussion unfolded across a series of interviews posted by FreightWaves, a logistics news outlet with fifty-three thousand followers on X, during its live coverage of the F3: Future of Freight Festival. The posts featured conversations with Urav Shah, founder and CEO of Evos, Danny Frisco of Lean Solutions Group, and Jay Singh, who each addressed different facets of how AI is reshaping freight operations.
In one segment, Frisco cited new research from the Everest Group finding that while sixty-nine percent of businesses have started on their AI journey—some past the pilot stage and into production—only nineteen percent have actually redesigned their workflows for AI.
"The issue that we've seen so far is about only nineteen percent of those businesses have actually redesigned their workflows for AI," Frisco said. "I think part of the issue is today AI is such a buzz word in the industry and there's so many point solutions out there that it demos really well and people jump into it a little too soon."
Frisco argued that companies need to build a foundation before deploying AI agents. "You have to get the data right. You have to get the SLPs right. You have to get the workforce management right. The re-skilling and upscaling of our employees, of your employees, of what we called experts in the loop," he said. "A lot of people just think, hey, we can plug it in, we can turn it on, we can get the agents going. And great, we've automated it and no more work to be done and I think that couldn't be farther from the truth so far."
The conversation also touched on the risks of relying on improvised systems. One segment addressed the use of Google Sheets as a dispatch tool. "I've actually seen a challenge where a customer was using a Google spreadsheet and one person was deleting a record and it was throwing everything in a tailspin," the speaker recounted. "That was not too long ago. That was one year ago and today that same, you know similar customer is working through a very creative flow of what a life of a load looks like."
The discussion described breaking down the life of a load—from booking through rate confirmation, tracking, document collection, and final payment—to identify what can be automated depending on the customer's needs. "We can automate the whole cycle," the speaker said. "There are many people in this room that could do that. But how do you actually need the customer where they are and what they want to automate is the critical portion."
Jay Singh, in a separate segment, addressed the spreadsheet problem directly, noting that running freight operations out of a Google Sheet works until a single deleted row throws an entire dispatch board into a tailspin. He described why modern logistics teams are moving away from such ad hoc tools toward more robust systems.
Urav Shah of Evos described the challenge of capturing "tribal knowledge"—the unwritten expertise that experienced dispatchers rely on—and teaching it to AI models. He explained how combining hard operational data with deep learning can replicate how a top dispatcher thinks, and he discussed the shift from legacy electronic data interchange and manual application programming interfaces to autonomous agent-to-agent communication.
Shah also showcased how AI can manage less-than-truckload bills across dozens of carrier portals, a task that previously required massive back-office teams manually hunting down billing discrepancies.
On the question of whether AI will ever completely eliminate human brokers and dispatchers, Frisco said automation takes over mundane repetitive work, freeing professionals to apply high-judgment critical thinking. "Humans are critical in the industry," he said. "Our industry is built on exceptions. I spent about ten years in logistics all on the carrier sales side. So I saw at first hand relationships are important. Problem solving is important. High judgment is important. And at the end of the day, customers and drivers sometimes just want a human. They want the empathy from the human."
He added that AI excels at low-level, tier-one tasks such as data entry, check calls, and document collection, freeing humans to work on claims, build relationships with carriers, and secure more capacity. "They will never not be a critical piece in our industry," Frisco said of human workers.
The posts and transcripts do not independently confirm the statistics cited from the Everest Group report or the specific incidents described by the speakers. Next News Network could not independently verify the claims made during the festival interviews.
Our Take
The freight industry's AI conversation is a useful corrective to the hype that dominates so much of the technology debate. While Wall Street and Silicon Valley promote AI as a plug-and-play solution that will eliminate back-office jobs and streamline operations overnight, the people actually running trucks are saying something far more grounded: the technology only works if you do the unglamorous work of fixing your data, retraining your workforce, and redesigning your processes first.
That only nineteen percent of companies have redesigned their workflows for AI—even as sixty-nine percent claim to be on an AI journey—should be a wake-up call. It suggests that many firms are buying tools without changing how they operate, a recipe for wasted money and unmet expectations. The lesson from the dispatcher whose Google Sheet was thrown into chaos by a single deleted row is that automation without robust systems is not progress; it is fragility dressed up as innovation.
Most importantly, the experts are not calling for the elimination of human workers. They are calling for the elevation of human judgment. In an industry built on exceptions and relationships, the empathy and problem-solving that only people can provide remain indispensable. AI can handle the repetitive tasks, but it cannot replace the dispatcher who knows which carrier will come through in a pinch or the broker who can calm a frustrated customer. That is a message worth hearing as the freight industry navigates its AI future.
