Survey: People don't trust AI, but they tell it everything (proton.me)

A recent survey by Proton reveals a paradox in user behavior towards AI chatbots: while two-thirds of participants have shared sensitive information—such as personal finances and mental health concerns—only a small fraction trust AI companies to protect their data. Despite the willingness to engage deeply with chatbots, trust in big tech remains low, with only 11% to 20% expressing high confidence in these companies' ability to safeguard privacy. Eamonn Maguire, Proton's director of AI engineering, highlights that AI systems understand user nuances far beyond mere keywords, indicating that users often disregard privacy concerns for the perceived convenience and lack of judgment in discussions. Key findings indicate that convenience drives users towards AI for sensitive topics, though friends and family remain the preferred confidants overall. Notably, privacy guarantees could significantly enhance user willingness to share information, with 48% expressing higher readiness to divulge details if conversations were assured not to train AI models. The findings underscore the urgent need for AI developers to prioritize trust and data protection to maintain user engagement, as enhancing privacy measures could determine the future trajectory of AI chatbot usage.