By Stuart Kerr, Technology Correspondent, LiveAIWire
Ford has rehired more than 300 veteran engineers after concluding that artificial intelligence alone could not replicate the quality judgement of the experienced staff it had let go. The admission, made by company executives this week, lands days after Ford topped the JD Power 2026 Initial Quality Study for the first time in sixteen years, a turnaround the automaker now attributes directly to bringing back the human expertise its AI systems were supposed to replace.
“Artificial intelligence is a fantastic tool, but it’s only as good as the information you use to train it,” Charles Poon, Ford’s vice president of vehicle hardware engineering, told reporters this week. “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.”
What Actually Went Wrong
Ford had expanded its use of AI across quality control in recent years, including rolling out 900 AI-powered cameras across its plants to catch defects and supply chain issues at the source, according to comments from chief operating officer Kumar Galhotra on an October earnings call. The expectation was that ingesting the company’s design requirements into AI systems would be sufficient to maintain product quality on its own.
That assumption did not hold. Poon told reporters the company had mistakenly believed that introducing AI and feeding it existing design requirements would automatically produce a high-quality product. The problem was not that the AI was fundamentally broken. It was that many of Ford’s most experienced engineers, the people whose decades of accumulated judgement the AI systems needed to learn from, had already left the company before that knowledge could be captured and used to train the tools meant to replace them.
The Rehiring and What It Reveals
Ford has spent the past three years bringing back what insiders refer to as “gray beard” engineers, technicians with decades of design and manufacturing experience who are now being used for two purposes simultaneously: catching the quality issues that automated systems were missing, and mentoring younger staff while helping rebuild and retrain the AI tools that fell short the first time.
“We recognised that for us to enhance some of our automation and machine learning and artificial intelligence tools we needed to ensure that they were trained by the most experienced individuals,” Poon said. The company’s press release marking its JD Power result was direct about the cause: “reaching best-in-class quality required a significant talent refresh.” That refresh involved not only the roughly 300 rehired engineers but also a turnover of senior leadership across engineering, supply chain, and manufacturing.
The Quality Numbers and the Asterisk
The results are measurable. Ford ranked first among mainstream automakers in JD Power’s 2026 Initial Quality Study, scoring 152 problems per 100 vehicles and placing the F-150, Mustang, and Super Duty at the top of their respective segments for the second consecutive year. CEO Jim Farley said falling warranty and recall costs were delivering “literally hundreds and hundreds of millions of dollars of a tailwind for Ford on cost.”
The turnaround has a significant qualifier attached. Ford remains the most recalled automaker in the United States, having issued 51 recalls in 2026 covering more than 11 million vehicles, more than double the next closest manufacturer. Galhotra has described the recall figures as a trailing measure of quality rather than a current one, arguing that as vehicles built under the revised engineering approach make up a larger share of the fleet, the recall numbers should improve correspondingly. Whether that prediction holds will not be clear for some time, since recalls by nature reflect defects in vehicles built one or more years earlier.
What This Means for You
Ford’s experience adds a concrete, named, on-the-record data point to a pattern that has been showing up across multiple industries: AI systems deployed to replace experienced human judgement frequently underperform unless that human judgement is actively used to train and supervise them, and removing experienced workers before their knowledge is captured creates a gap that the AI cannot independently fill. This is distinct from saying AI does not work. Ford is not abandoning its AI investment; it is restructuring how human expertise and AI tools work together, with veteran staff training and correcting the systems rather than being replaced by them outright.
For workers in roles where employers are weighing AI-driven automation, Ford’s public admission offers a useful data point: the value of accumulated, specific, hard-to-codify professional judgement does not disappear simply because an AI system has been trained on the available documentation. For organisations making similar bets, the lesson Ford is now telling reporters is one of timing and sequencing: capturing institutional knowledge before veteran staff leave, rather than after, determines whether an AI quality system actually works or simply automates the absence of judgement at scale.
LiveAIWire’s coverage of what the IMF’s AI job exposure figures actually mean for workers provides broader context on how AI is reshaping employment across sectors, a dynamic Ford’s experience illustrates concretely rather than theoretically.
About the Author
Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, emerging technology, and their impact on business, society, and everyday life. LiveAIWire publishes original AI journalism every weekday at liveaiwire.com.