A farmer destroys 10 hectares of crops after following AI advice

A 67-year-old farmer in Chuzhou, China, reportedly lost more than 10 hectares of sesame crops after relying on a chatbot for weed and pest-control guidance. He had used AI for farming advice for over a year. Early successes gradually made him trust its recommendations, and he followed a suggested mixture of chemicals without checking the instructions with an agricultural specialist or verifying them elsewhere. The weeds died, but so did the sesame plants. The account says one chemical was meant to be applied only to weeds, not sprayed across the crop, and the AI had failed to make that limitation clear.

The incident illustrates the risks of treating fast, confident AI responses as authoritative, especially when advice has real-world consequences. The farmer also overlooked the chatbot’s warning that generated text can contain errors. The point is not that AI should never be used: it can be useful, but its recommendations need independent verification. A second search or consultation with a qualified technician could have exposed the danger before the chemicals were applied. More broadly, repeated success with a tool can encourage people to stop questioning it, even though AI can produce faulty guidance; convenience should not replace critical judgment.