Ford brings back veteran engineers to fix AI quality gaps

Ford Motor Co. has brought back about 350 veteran engineers over the past three years after discovering that automated quality systems were not catching problems early enough. The company said the hires, promotions and returns include former Ford employees and engineers who had been working at suppliers, all part of a push to restore hands-on judgment to product development.

Charles Poon, Ford’s vice president of vehicle hardware engineering, said the company had made a basic mistake in assuming software could do the job on its own. Ford, he said, thought that simply introducing artificial intelligence and feeding in design requirements would be enough to produce a high-quality product. Instead, the automaker is leaning on experienced engineers to do work that AI struggled to handle: spotting failure points before parts reach the plant floor, mentoring younger staff and reprogramming the tools so they can flag defects earlier in development.

Kumar Galhotra, Ford’s chief operating officer, said those veteran engineers were “at the heart” of the company’s quality turnaround. Ford said the group has also helped improve the data collection and training processes behind its AI systems, a sign that the company has not abandoned automation but is trying to make it more useful by adding human experience and judgment. The company’s broader shift is from a “find-and-fix” mentality to one centered on preventing defects before they happen.

The move comes after years of costly recalls and warranty problems that hurt Ford’s reputation. Ford said it recorded lower year-over-year warranty costs in 2025, and the quality effort coincided with a strong result in J.D. Power’s 2026 U.S. Initial Quality Study, which measures problems per 100 vehicles in the first 90 days of ownership. Ford ranked first among mainstream brands, its first time leading that category in 16 years.

The result also underscored a wider lesson for manufacturers racing to automate production and engineering: AI can help, but it still needs the accumulated knowledge of people who have seen multiple product cycles, traced recurring failures and learned where design assumptions break down. Ford’s bet is that those “gray beard” engineers can make the software sharper, not replace it.