30% of AI Chatbots Share User Data, Reports Reveal

A new report from Surfshark reveals alarming privacy violations by AI chatbots, with 30% sharing user data with third parties, including data brokers. Google Gemini stands out as the most invasive, collecting 22 out of 35 possible data types, while other chatbots like ChatGPT and Jasper also engage in data tracking for advertising purposes. In Germany, as the federal election approaches, two AI chatbots—Wahl-O-Mat and Wahl.chat—aim to assist voters in making informed decisions, though concerns about bias have emerged with some chatbots favoring left-leaning parties. Wahl.chat, developed by Cambridge doctoral students, seeks to provide an objective overview of party policies to undecided voters. The integration of AI in government services is also highlighted, showcasing chatbots that enhance efficiency and reduce costs for caseworkers while improving citizen engagement. Overall, the intersection of AI and data privacy raises significant questions as chatbots become more prevalent in both personal and civic contexts.