AI in safety: what to consider before jumping in

November 5, 2025

AI is all the craze now, across the board. In trucking, some fleets are dipping their toes in AI for route optimization, driver assistance, predictive maintenance—you name it. But putting FOMO aside, is AI right for your safety program?

That’s the question many fleet managers are racking their brains with. In the September 2025 webinar Inside AI and Safety, CarriersEdge co-founder Mark Murrell unpacked the promises, pitfalls, and practical questions fleets should ask before adopting any AI-powered tools.

The bottom line? Not every fleet needs AI right now, and that’s okay.

Don’t adopt AI because it’s trendy

Start with the right question: Why now? Before adopting any new tool, especially one as hyped and misunderstood as AI, fleets need to ask themselves what problem they’re trying to solve.

Is it to reduce preventable incidents? Improve coaching consistency? Automate reporting? If the answer is “because everyone else is doing it,” pause right there.

Murrell compares this moment in tech to the early days of electricity before people really knew how apply it to their tools. “It’s like they’re wiring up their hammers and saws and electrocuting themselves in the process.” The potential was real, but it took time and plenty of missteps before they produced power drills and sanders.

A more recent example? The dot-com boom. Lots of experimentation, few proven solutions. And while AI might become a standard tool one day, right now it’s more about pilot programs than plug-and-play platforms.

That hype cycle around AI is self-perpetuating. Investors push companies to incorporate AI so they don’t look behind-the-times. Vendors, in turn, feel pressure to promote it in their products. Customers start asking about it all the time because everyone is talking about it. And that feeds the hype cycle. But hype isn’t a reason to act. Step back and look at it rationally.

Fleet size matters

Large carriers with 1,000+ trucks can afford experimenting with AI. They have the people and systems in place to manage false positives, massive data volume, and pilot testing.

Smaller fleets (50-200 trucks), however, may find that adopting AI prematurely creates more problems than it solves. Managing all the extra data, resolving false alerts, and adapting workflows may stretch already lean teams.

If you’re in the small to mid-sized category and haven’t touched AI yet, you’re not behind. You’re being smart.

Don’t replace human relationships

AI can process patterns, detect trends, and surface risks. But it can’t build trust.

Many fleets have tried to use AI tools, like dashcams or chatbots, to replace driver-facing interactions. That approach rarely works. Drivers already experience fatigue with in-cab alerts and automated safety nudges. If those alerts aren’t accurate, or if follow-up feels robotic, the tools quickly lose credibility.

“AI should never be the first point of contact with drivers,” Murrell says. “Use it to inform conversations, not replace them.”

If you try it, try it smart

If you’re ready to explore AI, approach it like any other safety investment: solve real problems, start small, and include the people it affects most.

• Include drivers from day one. When drivers feel ownership, they often become strong advocates. AI tools that provide better data and support safety coaching are generally well received, if drivers are included in the rollout. Fleets that simply drop the tools on their teams without explanation or input tend to meet resistance. But if drivers help shape how AI gets used, adoption is smoother, and results are stronger.
• Start with a business problem. Don’t chase the technology. Solve a real issue. Start by identifying a specific issue you want to solve, then look for tools that help with that. Once your goals are clear, it makes sense to talk to other fleets with similar challenges or lean on associations, safety groups, or your existing vendors to explore what’s worked elsewhere.
• Pick a champion. Designate someone in your team (ideally a curious driver or tech-inclined staffer) to explore tools and report back. This “AI champion” doesn’t have to be a driver, but it helps if they’re part of the driver community. When early adopters within the fleet are empowered to explore new systems and share their insights, the rest of the team is more likely to follow.
• Pilot small and build from there. Test one solution with one group. Gather input and adjust accordingly. If it works, expand gradually. The best results happen when AI rollouts are iterative, inclusive, and based on real-world feedback. Done right, even skeptical drivers can become advocates, and AI becomes a tool that enhances your safety culture.

Technology is liability if you ignore it

AI tools generate lots of data. And if you’re collecting it but not acting on it, you’re potentially exposing your fleet to liability.

Murrell warns, “Before, you could say, ‘We weren’t aware the driver was doing this.’ Now, with all the data, you have no excuse. If you’re not reviewing and acting on it, it looks like negligence.”

The lesson? Treat AI like a co-pilot, not a hands-free driver. Keep a human in the loop and review the insights regularly.

You’re not behind

Trucking isn’t late to AI. It’s as cautious as it should be.

“Eventually, AI will be like a power tool,” says Murrell. “It will become the nail gun, or the bandsaw, or the belt sander, but it won't replace all of the other tools.”

Until then, let the hype pass. Focus on what’s relevant and manageable. And when you do adopt AI, do it because it fits, not because it’s fashionable.