Ford realized that artificial intelligence requires human expertise
Artificial intelligence is increasingly being used in various fields, but companies are beginning to realize that this technology has a high cost and is prone to errors. Ford faced both problems while trying to use AI to improve vehicle quality, reduce warranty costs, and cut the number of recalls. Reality turned out to be more complex than the theoretical benefits promised.
During a recent conversation with journalists after Ford ranked first among mass brands in the JD Power Initial Quality Study, the company discussed how important AI has become to it but acknowledged its shortcomings. The unsuccessful implementation of the technology and the underestimation of experienced engineers initially even worsened quality.
When the experts left, the knowledge disappeared too
Ford’s Vice President of Vehicle Systems Engineering, Charles Poon, told The Verge that many experienced employees left the company before their knowledge could be transferred to artificial intelligence systems. To fill this gap, Ford had to hire, promote, and bring back more than 350 engineers who now train AI systems and improve data collection methods for AI tools.
Some of these engineers now mentor junior colleagues who previously could not handle maintaining vehicle quality. Poon noted:
It is our most experienced engineers who have the expertise to solve and detect problems before they enter the system.
He also acknowledged the company’s mistaken assumptions:
We mistakenly believed that simply implementing artificial intelligence and adjusting design requirements would result in a high-quality product.
Testing with AI is still important
Despite these mistakes, AI plays an important role in ensuring the quality of Ford vehicles. The company now uses over 100,000 artificial intelligence-based tests to check software systems and detect edge cases. If problems are found, changes to the software can be made quickly, even in the late stages of new model development.
Poon explained:
Because these tests are highly automated, even with a late software change, we can quickly run the entire validation process to ensure everything works perfectly before the car reaches the customer.
He added that Ford now treats software reliability as a separate strict discipline with clear metrics, similar to how it was previously applied only to hardware.
Transition from fixing to preventing
Solving quality issues also required a change in mindset. Poon told The Verge that Ford previously adhered to a “find and fix” philosophy, detecting defects after they appeared and eliminating them. Now the company aims to detect problems before they occur.
To achieve this goal, Ford’s software and digital teams are working more closely with engineering, manufacturing, and logistics divisions. The automaker also created a software quality assurance team of 40 employees dedicated exclusively to problem prevention.
Poon noted that Ford previously detected software errors in the late stages of development because it did not fully utilize the capabilities of rapid iteration. However, he emphasized that the company cannot adopt the approach typical of consumer electronics, “move fast and fix later,” since cars operate in a safety-critical environment where software must work correctly from the moment it reaches the customer.
Ford claims these changes are already yielding results. Although the company still has more recalls than any of its competitors in the US this year, warranty costs and recall frequency are declining.
This case illustrates an important lesson for the entire automotive industry: artificial intelligence is a powerful tool, but it cannot completely replace human experience and intuition. The return of experienced engineers to train AI systems shows that the most valuable knowledge is often stored in people’s minds, not in databases. Additionally, the transition from reactive “fixing errors” to proactive “preventing problems” requires not only technological changes but also a fundamental shift in corporate culture. For Ford, which has historically struggled with quality issues, this path may be the key to restoring consumer trust and reducing costs associated with warranty repairs and recalls.