Hertz Had a Good Idea. They Just Forgot the Governance.

The internet buzzed last week as reports of Hertz's AI misadventure hit the feeds. A customer returned a rental car, went about their business, and got an automated bill for $440. Curb rash. Detected by an AI-powered inspection system built by a company called UVeye, which uses cameras and machine learning to scan vehicles as they roll through the lot. No human looked at the car. No human reviewed the charge. No human was available to discuss.

Look, I get the appeal. Car rental return inspections are slow, inconsistent, and annoying for everyone involved. A walk-around with a bored employee holding a clipboard is not exactly a high-trust process. So Hertz partnered with UVeye to automate it, and on paper, this makes a ton of sense. Faster turnarounds. Consistent documentation. Fewer disputes about who scratched what. A genuinely good idea.

Here is where it gets interesting, though. Hertz's press release about the UVeye partnership said the system would "complement manual checks." UVeye's press materials talked about "safety and maintenance inspections." Neither company described what actually happened, which is that the system ran fully autonomous billing decisions with zero human review. The customer had no idea they were being assessed by a machine, no opportunity to contest any findings before being charged, and no clear path to reach a human who could look at the actual car.

The AI worked exactly as designed. It found damage, generated a charge, and sent the bill. From a pure technology standpoint, everything functioned. The problem is that nobody at Hertz appears to have sat down and asked the uncomfortable questions before turning it on. What happens when the system flags something the customer disagrees with? Who owns that decision? What do we tell customers about how the inspection actually works?

Three things would have made this a success story instead of a viral cautionary tale. First, honest communication. Tell customers when an AI system is inspecting the vehicle and explain what that means. People are surprisingly okay with automation when you're straight with them. Second, a path back to a human. Somewhere in the process, before a charge hits the credit card, a person needs to be reachable. Not a chatbot. Not a form. A person who can look at the photo and say, "Yeah, that's pre-existing," or "No, that's new." Third, someone at Hertz needed to own the risk decision. Not the vendor. Not the engineering team. Someone in the business who said I understand what this system does, I understand what can go wrong, and I'm accountable for how we deploy it.

That is what modern governance looks like. Not paperwork or checklists you hand to legal. It's the boring, necessary work of making sure good ideas don't blow up because someone forgot about the humans on the other side.

Hertz will fix this. They'll add a review step, a notification window, or a dispute button. But they'll do it reactively, after the bad press, viral posts, and damage to their brand. Companies that figure this out before the headlines are the ones that will actually capture the value that AI inspection systems can deliver. There is real value there; they just forgot to govern it.