AI isn’t working, so Ford is recalling 350 human engineers

There is one figure, however, that deserves attention: Ford itself remains the US car manufacturer with the highest number of recalls in 2026 (51 campaigns in the first six months). Regardless of the extent to which AI has influenced this record, the U-turn nevertheless appears to be an attempt to repair the damage. And this is where the news report ends and the part that makes the story interesting begins. Because Ford is not the first company to backtrack. In fact, it is merely the latest.

The most compelling precedent is that of Klarna. In 2024, the Swedish fintech company boasted of an AI assistant, built with OpenAI, capable of doing the work of seven hundred customer service staff. In early 2026, founder and CEO Sebastian Siemiatkowski told Bloomberg that this strategy was not the right approach and that the company was returning to hiring staff, because a customer must always be able to speak to a human being when they ask to do so. The gist of his change of heart, as reported by Bloomberg, is simple: they had gone too far. Of course, it should also be noted that the Swedish company itself has rejected the term ‘U-turn’, pointing out that it never completely phased out human support. But the direction has changed. And the case has become the benchmark for those in the sector looking at what happens when people are replaced all at once.

The same scenario played out again, only more spectacularly, at the fast-food restaurant. McDonald’s abandoned its experiment with automated drive-through ordering – developed in collaboration with IBM – following a series of blunders that went viral on social media, ranging from bacon being poured into ice cream to orders inflated with hundreds of croquettes.

Taco Bell, having processed more than two million orders via its automated voice system, is now rethinking the entire project following a number of high-profile incidents (such as the one involving a customer who ordered eighteen thousand glasses of water to crash the system and get through to a real-life operator).

The common thread running through all these cases is that automation has ended up relying on humans to rectify its mistakes.

And it is worth pointing out that these are not isolated incidents. According to data from the analytics firm Visier reported by Axios, the proportion of employees who are made redundant and then rehired by the same employer is on the rise – a sign that many companies are coming face to face with the limitations of artificial intelligence and are returning to investing in people.

However, a degree of caution is still warranted, and this applies to Ford as much as to others. Some of those job cuts attributed to AI were not actually driven by AI. Several research firms speak openly of ‘AI washing’ – that is, the practice of presenting redundancies decided for financial reasons as the result of the technology. In short, it is a way of shifting the blame for certain job cuts onto AI. After all, artificial intelligence isn’t going to take offence. At least not yet.