The Rise of AI Chatbot Leaks: How Users Unknowingly Share Sensitive Information

Abstract

While such AI chatbots like ChatGPT, Gemini, and Claude are increasingly becoming part of regular workflows and con- versations, privacy and data security concerns are amassing. Though the tools are immensely valuable to drive productivity, automation, and customer interactions, unknowing users are sharing close personal or business information during use. This study probes the prevalence of accidental data spillage by AI chatbots using the integration of a standardized survey of users alongside case studies predicated on real events. An 18–30 years’ old user born in the online world is the target user sought to track the usage behavior, type of information exchanged, interpretation of responses given by chatbots, and sensibility towards potential privacy threats. Moreover, well- known data breach case studies are analyzed to outline the manner in which misuse or misconfiguration of AI tools leads to gross privacy breaches. Implications of primary and secondary data indicate an acute lag in awareness among users, emphasizing the demand for more protection, policy measures, and AI tool digital literacy. The paper concludes by propounding practical suggestions to mitigate threats while unleashing the potential of generative AI.