Bumblebees solve novel puzzle, challenging assumptions about animal insight

Bumblebees have pulled off a feat that once seemed reserved for much larger-brained animals: they solved a completely novel object-manipulation puzzle without being taught the solution. In a study published in Science on June 4, 2026, researchers from the University of Oulu, the University of Helsinki and the University of Turku showed that buff-tailed bumblebees, Bombus terrestris, could move a ball beneath a blue artificial flower that had been relocated to the ceiling of a transparent arena, then climb onto the ball to reach a sugar reward.

The setup mattered as much as the result. The bees were first trained only on two separate ideas, that a blue flower signaled a reward and that the ball was a movable object. They were never shown the solution to the final task, yet many of them independently discovered it. Control experiments were designed to strip away easy explanations, and the researchers said the bees’ behavior looked goal-directed rather than like a simple response to visual feedback or reinforcement.

That is why the finding has drawn attention well beyond entomology. More than 70 percent of the bees figured out the task, a striking result for a species with a tiny brain and no previous experience with that kind of challenge. The work also echoes Wolfgang Köhler’s early chimpanzee experiments, long treated as a landmark case for insight and spontaneous problem-solving. Here, the same basic question was tested in an insect, and the answer suggests that flexible problem-solving is not confined to humans and other large-brained vertebrates.

The study adds to a growing body of evidence that bumblebees can do far more than once assumed. A 2024 line of work from the same research group found that bumblebees could cooperate on complex tasks and even wait for a partner before completing a shared goal. Another Science study in 2026 reported that they can form flexible abstract rhythm representations, a kind of cognitive processing previously demonstrated mainly in birds and mammals. Together, those findings point to a broader scientific shift: cognitive flexibility can emerge in very small brains, and the search for insight across species may need to look well beyond vertebrates.

The next step, researchers said, is to use slow-motion cameras and detailed video analysis to look for signs of an “aha” moment. That work could help separate true spontaneous insight from learned behavior, and deepen the debate over how intelligence should be measured across the animal kingdom.