GoKawiilDuckDuckGo survey finds users disclose deeply personal details to AI chatbots unknowingly
DuckDuckGo's new survey of nearly 2,000 U.S. adults found that almost a third of respondents have confided something to an AI chatbot they wouldn't tell friends, family or doctors, with the figure jumping to 56% among self-identified AI enthusiasts. The survey also found that most respondents didn't know chatbot conversations are typically used for training by default, and only a quarter knew such chats could be subpoenaed by law enforcement.
GoKawiil's interpretation of the reporting above, not reported fact.
The findings suggest a widening gap between how intimately people use AI chatbots and how little they understand about data retention, training use, and legal exposure. As AI companies market chatbots as trustworthy confidants, especially to parents who appear to disclose personal information at nearly double the rate of non-parents, the lack of transparency around data handling raises real privacy and legal risks for users.
- Nearly 1 in 3 users, and 56% of AI enthusiasts, share things with chatbots they wouldn't tell people close to them.
- Over half of respondents didn't know AI conversations are commonly used to train models by default.
- Parents confide sensitive information in AI chatbots at nearly twice the rate of non-parents.
Source: 9to5mac.com — Marcus Mendes, 2026-08-25
Published there as: “DuckDuckGo survey highlights how much personal information users share with AI”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.