Scientists at ETH Zurich have reported that chatbots can reveal users’ personal information.
Scientists at ETH Zurich noted that people share large amounts of information with bots without even realizing that it might be unsafe. If a person tells a chatbot that they took the morning tram, then the artificial intelligence can conclude that the user most likely lives in Europe, where trams are common.
Computer science professor Martin Vechev and his colleagues said that chatbots can collect and analyze user data by studying harmless dialogues. They can find out the user’s race, location, occupation, and more. This information can be used for targeted advertising or for fraud. About it writes Wired.