Ford Rehires 350 Veteran Engineers After AI Tools Trained on Bad Data Miss Quality Targets

Quick Facts

• Ford rehired 350 veteran engineers over three years to retrain AI quality tools and mentor younger staff.
• The move is expected to cut $1 billion in costs this year and pushed Ford to No. 1 in J.D. Power’s 2026 Initial Quality Study among mainstream brands.
• Ford VP Charles Poon said the company “mistakenly” believed AI alone would produce quality vehicles, but the tools were trained on incomplete data after experienced engineers left.

Ford Motor Co. has spent three years quietly undoing one of its own mistakes. The automaker rehired 350 veteran engineers, many of them former employees and others from suppliers, after its AI-powered quality systems failed to stop a cascade of defects that cost the company billions of dollars.

The effort appears to be working. Ford expects $1 billion in reduced costs this year. In J.D. Power’s 2026 Initial Quality Study, Ford ranked first among mainstream brands for the first time in 16 years, scoring 152 problems per 100 vehicles. That was a 41-problem improvement over the prior year, the largest year-over-year gain of any mainstream automaker. The F-150, Mustang, and Super Duty each won best-in-segment awards for the second straight year.

The story behind the turnaround is a lesson in what happens when companies remove the people who train the machines.

The Knowledge Gap

Ford VP of Vehicle Hardware Engineering Charles Poon told Bloomberg the problem was not the AI itself. “Artificial intelligence is a fantastic tool, but it’s only as good as the information you use to train it,” Poon said. When experienced engineers left before transferring their institutional knowledge, the AI tools were trained on incomplete inputs and amplified weak data rather than catching design flaws.

“Over prior years, we didn’t pay as much attention as we should have to the experience of our most knowledgeable engineers that have been with us through many product cycles,” Poon said. Teams spanning software, hardware, manufacturing, and supply chain had worked in isolation, reinforcing a reactive approach where defects surfaced late and were corrected under pressure.

What the Engineers Are Actually Doing

The rehired engineers are not returning to their old roles. Ford Chief Operating Officer Kumar Galhotra said they are at the heart of the quality turnaround, running mandatory troubleshooting meetings and reprogramming the AI tools that previously fell short. “We had been relying more and more on automated quality systems and not getting the desired results,” Galhotra said. “We brought back technical specialists and they hunt for failure points before a part ever reaches the plant floor.”

Ford has also added more than 100,000 new AI-powered tests to catch edge cases and stress software systems. A 40-member software quality assurance team now works to identify software issues before vehicles reach customers.

The Recall Problem Persists

The quality turnaround has not yet erased Ford’s recall record. The automaker leads all U.S. manufacturers in 2026, issuing 51 recalls so far covering more than 11 million vehicles, more than double the next-closest manufacturer. In 2025, Ford ran 153 separate recall campaigns affecting nearly 13 million vehicles, an average of one safety recall every 2.4 days. Ford expects those numbers to fall as upstream quality fixes take effect across newer vehicles.

A Broader Warning for Tech-Reliant Companies

Ford’s situation reflects a pattern playing out across Detroit. General Motors, Ford, and Stellantis have cut more than 20,000 U.S. salaried jobs combined, roughly 19% of their collective workforces, from recent employment peaks. Many of those cuts were tied to AI-driven efficiency initiatives.

Ford CEO Jim Farley has publicly predicted that AI will replace half of all white-collar workers in the U.S. Yet his own company’s quality crisis shows what happens when workforce reductions outpace AI readiness. Poon’s team now wants its systems to reflect not only computing power but also the practical engineering knowledge built up over decades of vehicle production.

For technology leaders, the Ford case offers a direct data point: AI systems are only as reliable as the training data behind them, and that data often lives inside the heads of people who are already out the door.

Read more: Ford rehires ‘gray beard’ engineers after AI falls short