Ford deployed 900 AI-powered inspection cameras and automated quality systems, then brought back 350 veteran specialists after the machines missed what experienced engineers catch. The company expects a billion dollars in reduced costs this year and just topped the JD Power Initial Quality Study among mainstream brands.

Charles Poon runs vehicle hardware engineering at Ford. Standing in front of journalists last week, he said the thing that executives at AI-forward companies almost never say out loud.

"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." — Charles Poon, VP of Vehicle Hardware Engineering at Ford (Bloomberg, June 25, 2026)

The confession came with a number attached. Ford has hired 350 veteran engineers, some of them former employees, others pulled from suppliers, after artificial intelligence and automated systems failed to deliver the quality levels the company expected. Ford calls them "gray beard" engineers.

For a company that set the all-time record for vehicle recalls in 2025, the stakes were not abstract.

The Automation Push Came First

Ford went into its quality overhaul the way most large companies have approached the last three years: with software. The automaker deployed 900 AI-powered cameras to detect quality issues on its lines and leaned into automated quality systems across engineering.

Chief Operating Officer Kumar Galhotra told journalists the company had been "relying more and more on automated quality systems," with disappointing results. Galhotra said teams across software, hardware, manufacturing, and supply chain often worked in isolation, which caught defects late and forced fixes under pressure.

The machines were watching. They were not understanding.

So Ford "brought back technical specialists," Galhotra said, veterans whose job is to "hunt for failure points before a part ever reaches the plant floor."

Experience Turned Out to Be the Missing Dataset

Poon's second admission cuts closer to the bone than his first.

"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." — Charles Poon (Bloomberg, via Motor1)

The gray beards are not replacing the AI. Ford is using them to do three things at once:

Hunt for failure points in parts and designs before anything reaches the plant floor, the pattern-matching work that comes from living through multiple product cycles
Train younger staff who never absorbed that institutional knowledge, because the people who held it had left for suppliers or retirement
Reprogram the AI tools that fell short, feeding the systems the judgment they could not learn from design requirements alone

That last item is the detail worth sitting with. The fix for underperforming AI was not more AI. It was the people whose knowledge the AI was supposed to encode.

The Turnaround Showed Up in the Numbers

The rehiring is producing results Ford is happy to publicize.

The automaker claimed the top spot among mainstream brands in the JD Power Initial Quality Study released last week. Ford anticipates the quality push will lead to $1 billion in reduced costs this year, largely through lower warranty and recall expenses. CEO Jim Farley said the rehired engineers are already "contributing to literally hundreds and hundreds of millions of dollars of a tailwind for Ford on cost."

The recall picture is improving too, with a caveat: Ford is on pace for fewer recall events in 2026 than its record-setting 2025, though Motor1 notes the number of vehicles affected may end up higher by year's end.

The Contradiction Ford Is Living With

Hold Ford's two public positions side by side.

Farley has said publicly that artificial intelligence "is going to replace literally half of all white-collar workers." Last week, his own executives stood in front of reporters explaining that Ford's quality recovery depends on 350 humans whose expertise the AI could not replicate.

Both statements can be true. Ford is not abandoning automation; the cameras stay, and the AI tools are being retrained rather than retired. Motor1's read on the arrangement is pointed: Ford brought back human workers, in part, to train their own replacements.

The tension is playing out across the industry, not just in Dearborn. Companies have spent 2026 citing AI to justify deep cuts, from Oracle's 30,000-job reduction delivered by 6 AM email to Microsoft's first voluntary buyout program in 51 years. At the same time, JPMorgan now tracks whether its 65,000 engineers use AI and factors it into performance reviews. The push to automate expertise and the scramble to recover it are happening simultaneously, sometimes inside the same company.

Ford just became the clearest data point yet on what the recovery side looks like.

The Bottom Line

Ford automated quality control, watched quality fall short, and paid to bring back the humans it had learned to live without. The company says the reversal is worth a billion dollars this year, which makes it one of the rare AI corrections with a public price tag attached.

The lesson for every engineering organization is the same one Ford's executives delivered in unusually plain language: AI trained on design requirements does not contain the judgment of the people who lived through the failures. Institutional knowledge does not compress into a prompt.

The gray beards are back in Dearborn, catching what 900 cameras missed. Their assignment, in part, is to teach the machines to stop needing them. Nobody at Ford is saying what happens when that works.