Some fleets are walking away from certain trucks — not because of fuel economy or service costs — but because the telematics systems behind them are not robust enough, Geotab founder and CEO Neil Cawse said during his opening keynote at Geotab Connect in Las Vegas.

The shift, he said, demonstrates a broader change in how fleets evaluate equipment, with telematics becoming an integral part of fleets’ operational infrastructure, rather than an add-on perk it once used to be. And artificial intelligence (AI) is now becoming the layer that determines how effectively that infrastructure performs.

Cawse described AI as “one of the biggest transformations hitting us since the Industrial Revolution.” He said that while telematics systems captured much valuable information for years, fleets often relied on manual reviews and after-the-fact reporting. AI changes the equation by moving telematics from historical tracking toward predictive decision-making.

“The shift starting is this: AI isn’t assisting us anymore. It’s doing the work, and that’s the new reality,” Cawse said.

Instead of simply showing where a vehicle has been, AI systems can now analyze millions of data points to identify patterns, flag emerging risks and recommend adjustments before a breakdown or collision occurs.

This is why Cawse said Geotab’s goal is skilling up its AI to be “the world’s best fleet and safety manager.”

“This means you will have a trusted partner reporting to you, that listens intently and follow through on your needs. So just watch that come out over the next year,” he teased, saying focusing the company’s focus areas are operations, maintenance and safety aspects of fleet management.

AI needs coaching, too

And while AI-powered systems now shift toward doing the work — like auto‑prioritizing maintenance and work orders based on risk, coaching drivers, and answering operational questions — Cawse cautioned against blind reliance on AI.

“AI is incredibly smart, but smart is not always the same as reliable. The old fear was hallucination. That’s improving fast, the bigger risk is now different — AI can be too agreeable. It often tells you what you want to hear when the context is missing. It guesses instead of asking,” he said. “Give it clear context. Ask it to show you assumptions, push it to say what it doesn’t know, and this is how you turn AI into a true business tool.”

Cawse also added that AI’s effectiveness depends on the data it is being fed, saying that high-quality data must be digital, accessible and usable.

“AI is only as good as the data that it can access. Telematics is a critical part of this, but it’s only one part.”

For example, maintenance logs trapped in PDFs, inspection reports stored on paper, or compliance records in disconnected systems restrict AI’s ability to produce meaningful insights.

Cawse also talked about the concept of digital twins — digital representations of vehicles and their operating environments — to provide AI with the contextual awareness it needs.

“You can’t manage what you don’t measure,” he said. Assessing driver behavior, for example, requires understanding what was happening around the vehicle. Route optimization depends on real-time traffic conditions and, in the case of electric trucks, battery range.

Having all this data connected and integrated allows for it to be measured, recorded, interpreted and acted upon – all by AI and all within seconds.

Adapt or die

However, when talking about AI and its rapid growth in capabilities, Cawse said the shift has already crossed a threshold he did not expect to reach so quickly.

“Right now, there are very few developers in the world that can write better code than AI,” he told the attendees in the audience. While humans may offer deeper context around specific problems, he argued that AI is producing higher-quality code “much, much faster.”

“So, one sacred profession is being turned on its head in every tech company, including ours. So what is this? AI is magic. It’s a bubble. It’s quicksand, all at the same time, depending on where you look.”

He said many software developers aren’t writing software anymore — the AI is – adding developer teams now manage the AI teams.

Cawse also added that artificial intelligence is reshaping how software-as-a-service companies operate, lowering barriers to building internal tools and even customer-facing functions.

Legal, marketing, administration, training, video production and customer support functions at Geotab are already being restructured around AI-first workflows, he said.

“Let me give you one example. Jonathan, one of our lawyers, used the AI tools to build his own legal application, and because of that, avoided us buying another SaaS application. Yes, a lawyer who doesn’t code.”

But Cawse acknowledged the transition is not seamless. “We have a human change control problem,” he said of organizations struggling to adapt as quickly as the technology evolves.

He also acknowledged the labor implications.

“Do I worry about some of our teammates’ jobs? If I’m brutally honest? Yes. But this is not a Geotab problem. This is all of our problems. This is taking jobs and will continue to do so, but we need to move with the times and adapt or die,” he said. “What a scary and exciting time to be alive. But I’d rather be in the room figuring out how to lead through this than watching it happen from the sidelines.”

He added that leadership, not resistance, will be key to navigating towards an AI-first future. He said every leadership skill that fleet mangers have honed over the past 20 years will become an advantage with AI.

“You’re not just using AI. You’re leading it. You define the problem, coach the process, challenge the output and raise the standard. When you lead AI the way you lead people, the outcomes will be exceptional.”