The logistics industry doesn’t like AI, according to new statistics from Tech.co’s monthly industry surveys.
Across a six-month period, Tech.co found that the amount of logistics professionals with a positive outlook on the future of AI fell from 30% to just 11%.
The reasons behind that shift are complex, ranging from job insecurity to poor implementation and results. So, we spoke to a wide swath of fleet management experts to learn more:
Key Takeaways
- 16% of fleet professionals feel “very negative” about AI in March 2026, up from just 7% six months earlier.
- Also, 65% have no AI exposure at work, up from just 31% six months earlier.
- To work, AI tools must be properly integrated with a focus on interoperability and human oversight.
- AI use within logistics may become more targeted, as companies avoid generic applications.
Positive Views of Logistics AI Fell by Two-Thirds in Six Months
When asked how they would describe their “personal outlook on the future of AI,” 30% of fleet professionals told Tech.co it was positive, in a September 2025 survey. By March 2026, only 11% said the same.
During that same time period, negative sentiments toward AI increased. 14% have “somewhat negative” views of logistics AI in March 2026, up from 9% last September, while 16% feel “very negative,” up from just 7% six months prior.
We also charted an increase in the amount of logistics professionals who have little-to-no exposure to AI. In September 2025, about one third (31%) had no AI exposure in their work, and in March 2026, two thirds (65%) have no AI exposure.
So, what’s behind the shift? We checked in with a number of industry experts for some additional boots-on-the-ground insight. Here are the biggest takeaways.
Quick Turnaround Times Make a Big Difference
Those working in the logistics business don’t want a pie-in-the-sky benefit that takes months to arrive, if it ever does. They’re facing tough financial pressure today, data shows, so AI tools that can deliver short-term benefits stand a better chance of finding a market.
That’s according to Dmytro Negodiuk, fractional AI officer at Negodiuk AI Consulting, who argues that logistics pros “quietly use [AI] for the small things — inbound calls, quote generation, follow-up — where payback is measurable inside 30 days”.
If AI use has shifted to these small-scale examples, which are invisible to anyone not in the room when it happens, this might explain the growing number of people with no AI exposure. They simply aren’t aware of AI use, even if it’s happening at their business.
Lack of Proper Integration Makes AI Worse Than Nothing
One big reason the pros are cooling on the idea of adding AI to workflows: too many operations are adding it thoughtlessly. It’s easy to throw a layer of AI chatbots over existing tools, but all it does is add an interface to the process that an operation was already doing before.
Founder and managing director of Pace Technology UK LTD Steve Kealey brings his experience in pharmaceutical logistics to the table when discussing this issue:
“We have seen operators move away from AI tools because earlier systems slowed [the process] down — too many dashboards, too many steps, and just not enough ability to get to a clear answer. I don’t think the issue is AI itself. It is whether it actually reduces friction in the operation. In this sector, if it does not help you get to the truth faster, it is not useful, no matter how advanced it looks on paper.”
– Steve Kealey
Rich Pleeth, co-founder at Finmile, says the same: “A lot of tools promised optimisation and delivered dashboards. They show you what’s wrong but still rely on people to fix it, so the workload doesn’t actually change.”
Key element: Oversight
So, how can your operation correctly investigate AI tools? Amy Dean, vice president of operations at SC Codeworks, emphasizes the value of human-led monitoring of AI output.
“AI can be incredibly powerful, but it can also be dangerous if it’s used in place of people, instead of alongside them. A lot of the negative sentiment we’re seeing comes from how quickly it’s being adopted without enough oversight. In logistics, where decisions have real operational consequences, you can’t afford to rely on something that isn’t grounded in human context and experience. People shouldn’t be looking to AI to do their job for them.”
– Amy Dean
This reaffirms what experts have previously told us: The sector wants humans to be held accountable for decisions.
Key element: Interoperability
Similarly, Nithin Mummaneni, CEO of Infinity Loop, says early enterprise deployments of AI agents have exposed a major operational gap.
Business workflows span multiple systems, Mummaneni explains: collaboration tools, CRMs, procurement platforms, and document repositories.
If AI agents move between these environments, they face plenty of failure risks, since they’ll need to adjust to new APIs or permissions. Humans can handle the interoperability better than AI tools, so they’re a great “middleware layer.”
“Companies are increasingly evaluating AI tools based on open API readiness and their ability to operate across existing systems rather than creating another silo. For areas like vendor contracts and negotiations, where data often sits across multiple platforms, AI only delivers value if it can integrate seamlessly into the broader enterprise workflow.”
– Nithin Mummaneni
Avoid the Generic
To work, predictive AI must understand the full shipment lifecycle, not just one element.
“For years, ecommerce logistics has been stitched together across too many vendors, leaving merchants drowning in disconnected tools and inaccessible data. The negative sentiment about AI often traces back to one root cause: AI was too generic,” says Kelly Vincent, chief product officer at ShipStation.
Ben Peters, CEO and co-founder of Cogna, agrees. “Most of what gets sold to these firms as ‘AI’ is generic software pretending to be something more advanced,” Peters says. “It doesn’t fit operator workflows, it doesn’t connect to the right data sets. So they don’t use it.”
Helping Humans — Not Replacing Them
“There’s a shift in how AI is being applied,” says Arun Samuga, chief innovation officer at Elemica. “Less focus on fully automating decisions, more on improving how teams respond — humans in the decision loop, not just in the training loop.”
Some examples of great AI use from Samuga include:
- Prioritizing which orders to expedite when multiple shipments are delayed and capacity is constrained
- Identifying inbound materials that will delay production schedules or create inventory gaps
- Extracting order details from emails or PDFs to onboard new customers or suppliers faster
That doesn’t stop fears of being replaced, however: Tech.co’s March 2026 industry survey found a 13 percentage-point rise in the belief that “AI will definitely lead to fewer job opportunities” alongside a 17 percentage-point drop in the belief that AI will definitely lead to more job opportunities.
So far, AI tools can support human decision-making but can’t replace it. AI tools can optimize fleets by sending alerts for maintenance issues or harsh driving events, but humans will still need to respond to those alerts.
Dodging the Hype by Limiting AI Usage
Artsiom Kozak, head of AI technical expertise at Innowise, notes that “tons of promises and overhyped expectations have been put on AI,” and the logistics business is no exception.
“At least right now, it cannot yet replace a dispatcher who knows their drivers personally — their fatigue levels, their quirks, whether they’ll stay an extra hour for a good parking spot. AI cannot reliably understand local constraints that aren’t in the data: a closed bridge, an aggressive security guard at a depot entrance, or a driver who always turns left just because that’s their habit.”
– Artsiom Kozak
AI isn’t a fix-all solution, and treating it like one can mask problems rather than address them.
Case study: Healthcare shipping firm Mercury
For some companies, AI use is strictly limited. Josh Medow, CEO at healthcare shipping firm Mercury, told me about his team’s “deliberately cautious approach to AI implementation”.
“Good logistics centers on navigating exceptions, and AI tools don’t have the capacity to do that. AI can help train team members, make suggestions, and fill out forms, but it cannot navigate real-world exceptions, like a shipment stuck at customs or a driver who got into a car accident. We didn’t want to create this risk for our customers for the simple goal of reducing costs or moving faster.”
– Josh Medow
Mercury controls its AI training, with Medow personally in charge of updating its knowledge source weekly. Once their operations team has tested AI functions, they may eventually roll them out more broadly.
“For us, responsibility in AI means using it to reduce pain points, not to replace the judgment our critical healthcare shipments require,” says Medow.
Some AI Logistics Tools Are Seeing Increasing Use
One last finding from Tech.co’s recent industry survey serves to underscore the point that many experts we spoke with made: AI isn’t going away. Our monthly surveys found that AI tools, like fraud detection systems, last mile delivery, and autonomous vehicles, have seen an increase since July 2025.
These tools can trigger negative reactions in logistics employees, either from job-security fears (autonomous vehicles might replace them) or because they’re unfamiliar with the functions (fraud prevention systems run in the background as “invisible” AI).
However, the operations that can address specific logistics challenges with targeted, human-verified AI tools can still streamline operations and make a difference on the frontline.