A new study highlights a hidden challenge of using AI in medical imaging research — the phenomenon of highly accurate yet potentially misleading results known as "shortcut learning." The researchers analyzed thousands of knee X-rays and found that AI models can "predict" unrelated and implausible traits such as whether patients abstained from eating refried beans or beer. While these predictions have no medical basis, the models achieved high levels of accuracy by exploiting subtle and unintended patterns in the data.