Ford Motor Co. (NYSE: F) spent years betting that artificial intelligence could pick up where its veteran engineers left off. And then it couldn’t.
The Detroit automaker acknowledged this week it had to bring in 350 experienced engineers after its automated quality-control systems fell short — a concession timed to Ford’s best performance in JD Power’s US Initial Quality Study in 16 years.
Charles Poon, Ford’s vice president of vehicle hardware engineering, told reporters that AI is “only as good as the information you use to train it” — experienced workers left before they could encode their institutional knowledge in Ford’s training data, leaving its automated systems to reinforce bad assumptions rather than surface the defects they were built to find.
The company has cut roughly 5,300 salaried roles since its 2020 headcount peak — part of a Detroit-wide pullback that erased more than 20,000 white-collar positions across the industry. Ford brought in the 350 engineers — a mix of rehires, new hires, and internal promotions — over three years to mentor junior staff, rebuild data pipelines, and recalibrate the AI systems originally intended to make them redundant.
Sponsored · Selkirk Copper Mines
The company also assembled a dedicated 40-person software quality-assurance team and deployed more than 100,000 AI-powered automated tests to flag edge cases late in development.
Ford climbed from No. 15 among mainstream brands in 2023 to No. 1 in JD Power’s 2026 US Initial Quality Study, scoring 152 problems per 100 vehicles against an industry average of 175 — the largest single-year improvement of any mainstream brand in the study.
Seven of Ford’s 10 tested models placed in the top three of their respective segments, the highest share of any automaker. The F-150, Mustang, and Super Duty each took best-in-segment for the second consecutive year. Ford anticipates the broader quality push will cut costs by $1 billion in 2026.
The Initial Quality Study measures only the first 90 days of ownership; in JD Power’s separate Vehicle Dependability Study — based on three years with the vehicle — Ford ranked 19th. The automaker has also issued 51 recalls in 2026 covering more than 11 million vehicles, more than double any other US manufacturer.
Chief Operating Officer Kumar Galhotra argued that recalls reflect older model quality and should improve as newer vehicles reach consumers.
CEO Jim Farley previously predicted AI would replace “literally half of all white-collar workers” in the US — a forecast that hasn’t aged well for Ford, at least not yet. Ford still runs on AI. What changed is the premise — machines need the experts, not the other way around.
