Ford Bet on AI Over Its Veteran Engineers, Then Rehired 350 of Them

Ford spent years leaning on artificial intelligence to fix quality problems that, by the company’s own account, have cost it billions: automated quality systems and hundreds of AI-powered cameras watching for defects on the line. The results fell short. So over the past three years, Bloomberg reported in June 2026, Ford hired 350 veteran engineers, some of them former employees and others pulled from suppliers, to fix what the automation could not. Inside the company they are called the "gray beards."

Charles Poon, Ford’s vice president of vehicle hardware engineering, described the original mistake to reporters with unusual candor: "Mistakenly we thought that by just introducing artificial intelligence and ingesting the design requirements that we had, that that would produce a high-quality product."

What Ford Got Wrong

The assumption underneath Ford’s push is one a lot of companies are making right now: that expertise can be captured by feeding an AI the written-down version of a job. Ingest the specs, the requirements, and the documented standards, and the system should be able to do what the experts did.

What Ford found is that much of what an experienced engineer knows was never written down. It lives in judgment built over many product cycles, in the intuition for which part is likely to fail and the pattern recognition that comes from seeing the same category of problem a hundred times. The AI had the documented requirements. The defects that mattered were being caught by the undocumented experience, and by the time Ford understood that, some of the people who carried it were gone. "Over prior years, we didn’t pay as much attention as we should have to the experience of our most knowledgeable engineers who have been with us through many product cycles," Poon said.

What the Recovery Looked Like

Kumar Galhotra, Ford’s chief operating officer, told reporters the company had been "relying more and more on automated quality systems" with disappointing results. The returning veterans were not brought back to replace those systems. According to Galhotra, they reprogrammed the AI tools to catch glitches earlier, now run mandatory quality meetings, and hunt for failure points before a part ever reaches the plant floor. The experienced engineers define what good looks like and teach the system; the AI then applies that judgment at scale.

The numbers suggest it worked. In the J.D. Power Initial Quality Study released June 25, 2026, Ford ranked as the top mainstream brand, up from tenth a year earlier, with only Porsche and Genesis scoring higher overall and the F-150, Super Duty, and Mustang each leading their segments. CEO Jim Farley put a dollar shape on it on Bloomberg TV: "We’re seeing our warranty coverages come down. We’re seeing our recall costs come down. These are all contributing to literally hundreds and hundreds of millions of dollars of a tailwind for Ford on cost."

That tailwind figure is worth reading in reverse. The same warranty, recall, and rework costs that are now falling were the price of running quality on automation that had never absorbed the experts’ judgment, paid for years before the correction, on top of whatever premium it took to bring 350 veterans back.

Before You Automate Expert Judgment

Ask how much of the skill is actually in the manuals. Some work is fully captured in written procedure, and AI can learn it from those documents. High-consequence expert judgment usually is not. If the most valuable part of a role lives in experience rather than documentation, the documentation alone cannot train a replacement for it.

Capture the expertise while you still employ it. Ford’s sharpest regret was that knowledgeable engineers left before what they knew could be transferred anywhere. If you intend to automate expert work, the experts have to be involved in building and training the system while they are still on payroll, and they should stay involved afterward to correct it.

Price the cleanup scenario, not just the savings. Headcount savings from automating experts show up immediately. The cost of being wrong arrives later as warranty claims, recalls, brand damage, and the premium to rehire, and at Ford it was large enough that reversing it now measures in hundreds of millions of dollars a year.

This case names the boundary we watch most carefully with clients: the line between work AI can take over and expertise AI can only support. The strongest deployments capture what your best people know and use AI to apply it consistently at scale. Ford’s three-year detour is what it costs to find that line the other way around.