The Very Real Pitfalls of Artificial Intelligence

March always shows up with the same promises: more light, milder temperatures... and guaranteed disappointment. You break out the sunglasses, slip on a patch of ice, and remember that winter hasn't had its last word yet. So rather than fighting March, you might as well welcome it with a sense of humor. With that in mind, this article looks at three moments when artificial intelligence made mistakes that led to comical consequences. Because, just like the month of March, this revolutionary tool can also fall short of our expectations.

Ski slopes as a road

We can all agree on one thing: GPS is now an essential tool, especially when traveling in a country or region you don't know. Its main advantage is to guide us efficiently by calculating the fastest or shortest route to our destination. However, this pursuit of optimization is not without risk, especially when artificial intelligence enters the picture... and gets it wrong.

In fact, in 2025, a person following their GPS directions to find their way along mountain roads ended up stuck in the snow. To save them time, the navigation system had suggested an alternative route that was supposed to be faster. As it turned out, this "road" was actually a ski slope. The result: the car got stuck, perfectly illustrating the limits of technology.

Chicken nuggets ordered by artificial intelligence

In 2019, McDonald's launched a project to bring artificial intelligence into drive-thru order taking. The main objective of this initiative was to reduce the human resources needed for taking orders while cutting wait times for customers at the drive-thru. However, the project was ultimately abandoned in 2024, following several amusing incidents that occurred during its use.

Indeed, many customers reported unusual errors in their orders. For example, one customer who simply wanted to order an ice cream ended up with their cone, several boxes of McNuggets, and various sauces added to their bill. Although the amount could be corrected afterward, the situation remains amusing to say the least. Faced with the repetition of this kind of incident, McDonald's decided to remove AI from the order-taking process, judging that the results did not meet the company's expectations and quality standards.

When AI invents case law

A final story takes place this time in a courtroom. As a rule, every good lawyer spends many hours on legal research in order to build a solid, convincing argument to effectively defend their client before a judge. This work demands rigor, patience, and precision, because every argument must rely on real, verifiable prior rulings. However, this research is often long and complex, which can sometimes create the temptation to take shortcuts.

That is precisely what happened to a lawyer in the State of New York. In order to strengthen his case, he turned to artificial intelligence to find court rulings similar to his case to support his argument. However, after submitting his file to the court, it turned out that the cited cases simply did not exist. The artificial intelligence had in fact fabricated fictitious court rulings, presented very credibly but entirely made up.

This situation rendered the lawyer's argument inadmissible in the judge's eyes and seriously damaged his professional credibility. The lawyer was severely reprimanded and sanctioned by the court for failing to verify the authenticity of his sources. This incident quickly became a striking example of the limits of artificial intelligence and the importance of human judgment, particularly in fields where rigor and accuracy are essential.

The lesson to take away

From all these stories, a clear lesson emerges. Although artificial intelligence represents a major advance and is revolutionizing the world of work, it remains imperfect. Like an intern in a learning phase, it requires constant supervision as well as rigorous validation of its work before its results are deployed or integrated into organizational processes. Human judgment therefore remains essential to govern its use and reap its full benefits, while avoiding mistakes and excesses.

Did you know? The MMO team is here to support you with your various projects, including the implementation and optimization of artificial intelligence within your work processes. Whether you need advice, support, or personalized assistance, do not hesitate to contact us: we will be happy to help you harness the full potential of AI, safely and effectively.