Ford Brings Back 350 Veteran Engineers After AI Quality Control Failure

Ford has admitted it leaned too heavily on artificial intelligence for vehicle quality control, and spent three years cleaning up the consequences of that decision. The company revealed that during that time it rehired 350 experienced engineers whose expertise the automated systems had been missing.

Ford's story is one of the first such open admissions by a major automaker that AI-driven automation failed in areas where experienced people used to work. The company isn't backing away from AI, but it has rebuilt its approach to how and where the system should support engineers rather than replace them.

What Went Wrong

Ford, like other Detroit automakers, sharply cut engineering headcount in recent years, betting that AI tools would take over part of the inspection and design work. Since 2020 the company has eliminated 5,300 white-collar positions, and the three largest Detroit automakers together have laid off roughly 20,000 office workers.

The problem was that many of the most experienced engineers left the company before their knowledge was transferred to the automated systems in any meaningful way. As a result, instead of catching design flaws, the AI tools reproduced and amplified the weak input data they had been trained on.

We mistakenly believed that simply introducing artificial intelligence and feeding it our design requirements would be enough to get a high-quality product - Charles Poon, Ford vice president of vehicle hardware engineering

The Veterans Return

The solution turned out to be bringing back so-called gray beard engineers, longtime staffers who had left or been laid off in earlier rounds of cuts. Their job isn't to replace AI, though, but to mentor younger employees and rebuild the machine learning systems so they actually catch design flaws.

Artificial intelligence is a fantastic tool, but it's only as good as the data it was trained on - Charles Poon, Ford vice president of vehicle hardware engineering

Ford stresses that AI remains a core part of the company's strategy. The returning engineers weren't tasked with rejecting automation, but with enriching it with the knowledge missing from the training data and building oversight mechanisms for the decisions made by the quality control systems.

Results and Numbers

The company added more than 100,000 automated AI-based tests to its production process and created a 40-person team responsible for software quality. Combining human experience with expanded automation translated into a measurable effect on industry rankings and warranty repair costs.

Ford took first place among mass-market brands in the prestigious JD Power Initial Quality Survey, for the first time in sixteen years. CEO Jim Farley said the drop in warranty costs and service actions translated into savings running into the hundreds of millions of dollars a year.

For Polish manufacturing companies considering deploying AI in quality control or design, Ford's case is a practical warning rather than an ideological one. Automation that doesn't preserve expert knowledge within the organization, whether as training data or documented procedures, can increase the risk of errors rather than reduce it, especially when experienced employees leave faster than the company can digitize their knowledge.

Sources: Forbes (forbes.com), Yahoo Finance (finance.yahoo.com), Motor1 (motor1.com)