I asked an AI chatbot for my data. I didn't expect a psychological profile (thelocalstack.eu)

In a revealing exploration, a premium AI chatbot user requested their stored data, ultimately uncovering a detailed psychological profile generated by the service. The analysis highlighted significant differences in data retention strategies among major AI platforms. OpenAI's ChatGPT amassed vast amounts of raw data, including text, voice recordings, and images, emphasizing a profiling approach based on extensive storage. In contrast, Anthropic's Claude offered a more curated insight, distilling user interactions into structured psychological assessments that mapped emotional states and behavioral patterns. Meanwhile, Perplexity generated robust insights through search queries, documenting specific user interests and behaviors without the depth found in other services. This case underscores critical implications for the AI/ML community, as users often unknowingly trade personal insights for convenience and support. The data reveals the potential for targeted manipulation and behavioral prediction based on users' vulnerabilities. As concerns about data privacy grow, the story also illustrates the importance of informed consent and user rights, such as utilizing GDPR regulations to understand what personal information is collected and how it’s utilized. The narrative serves as a reminder of the intricacies of AI interactions, prompting users to reevaluate their engagement with these tools and be aware of what they inadvertently share.