AI models get mushroom ID dangerously wrong. Don't eat what they tell you to

Key points

• Piotr Migdał, a software engineer at AI analytics firm Quesma, tested 16 AI models on 1,040 mushroom photos across 55 species in a study published in May 2025.
• The best-performing model, Google's Gemini 2.0 Flash, correctly identified a mushroom on its first guess only 65 percent of the time.
• The death cap, one of the world's most lethal mushrooms, was mistakenly identified as edible by AI models 16 percent of the time.
• The worst model, Qwen 3.8-27b (an AI from Chinese tech firm Alibaba), labelled poisonous mushrooms as edible 36 percent of the time.
• Migdał's verdict, shared with The Register, is blunt: never eat a mushroom because an AI said it was safe.

Piotr Migdał had a simple question. Could a modern AI chatbot, the kind millions of people use every day to answer questions, reliably tell a safe mushroom from a deadly one? He spent weeks finding out. The answer should stop you opening your camera app next time you spot a fungi in the woods.

Migdał, a founding engineer at the AI firm Quesma, built a test set of 1,040 photographs covering 55 mushroom species, drawing on a Danish mushroom library and records of safe and deadly species from Poland and beyond. He ran every photo through 16 of the leading AI image-recognition models, asking each one to name the most likely species and its four next-best guesses.

How badly did the AI models actually do?

Badly enough to worry. The top performer, Google's Gemini 2.0 Flash, got the species right on its first answer just 65 percent of the time. That sounds passable until you remember the other 35 percent of the time it was wrong, and wrong answers in mushroom hunting can be fatal.

At the bottom of the table sat Qwen 3.8-27b, a model built by Alibaba. It named the correct species first only 13 percent of the time.

The errors were not spread randomly across harmless mix-ups. Migdał found that the models kept making the exact mistakes that experienced foragers are warned about for decades. A deadly webcap, which kills by destroying the kidneys over several days, was called a chanterelle, the same confusion that has killed real people. The death cap, responsible for the majority of fatal mushroom poisonings worldwide, was called edible 16 percent of the time. The fool's funnel was misidentified as safe 48 percent of the time.

Should you use AI to check a mushroom you find?

No. Migdał is direct about it: "Do not eat a mushroom because AI told you it is safe."

One model, Muse-spark-1.2 from Meta, posted a deceptively low false-positive rate of eight percent, meaning it rarely called a poisonous mushroom edible. The catch: it simply refused to guess most of the time. That is actually the right call. Refusing to answer is safer than confident fiction.

Migdał told The Register he was genuinely impressed that general-purpose AI models could identify most mushrooms at all. But "most is not enough," he said, "as errors could have dire consequences."

Experienced human foragers follow a practical rule: stick to species with no dangerous look-alikes until you really know what you are doing. That rule applies even more firmly to AI. If the machine is wrong one time in three on the easy cases, it has no business advising you on the hard ones.

Common questions

Could a specialist mushroom app do better than a general chatbot?

Possibly, though Migdał's study tested general AI models rather than dedicated foraging apps. A model trained specifically on fungi and regional species differences might improve accuracy, but no app has been independently verified as safe enough to stake your health on.

What should you actually do if you find a mushroom you want to eat?

Go with an experienced human forager who knows the local species, or join a local mycology (mushroom-study) club. Failing that, stick to varieties sold in shops, where the identification work is already done.