When can we trust the results we get from AI, and when is learning impossible? Researchers have shown that there are some problems that even the most powerful AI cannot reliably solve, no matter how much data it is given.

The researchers, from the University of Cambridge and the University of California, Santa Barbara, designed “adversarial” mathematical systems to fool any AI algorithm. Like ethical hackers stress-testing a network’s security, these adversarial systems were designed to map out exactly where and why AI prediction breaks down.

Many real-world systems—like those in the oceans, the human brain or robotics—are too complex to describe neatly with equations, so researchers often learn how they behave by using machine learning. But these AI methods don’t always work well, returning unreliable results or poor predictions.