MENLO PARK, USA – Recent reports indicate that conversations and prompts submitted to Meta AI are appearing on a public "Discover" feed, raising significant privacy concerns and questions about whether users fully comprehend the public nature of their interactions. While Meta clarifies that content is only public if users explicitly choose to share it, a considerable number of highly personal and sometimes sensitive chats are finding their way into public view, suggesting widespread user confusion.

Meta's standalone AI chatbot app, launched in April, includes a "Share" button that, when tapped, leads to a "Post to feed" option. However, users appear to be misunderstanding this functionality, believing they are saving conversations privately or sharing them with specific contacts, rather than with the entire public. This has resulted in the "Discover" feed being populated with a stream of awkward, revealing, and even alarming conversations, including personal queries, photos, and audio clips.

Cybersecurity experts and privacy advocates are voicing strong concerns. They point out that in many cases, these public posts are easily traceable back to users' social media accounts through visible usernames and profile pictures, creating significant privacy and safety risks. Examples found on the feed include users discussing medical conditions, relationship problems, and even attempting to self-diagnose ailments.

While Meta has recently added new warning pop-ups to make the sharing process clearer, critics argue that the app's interface still employs "dark patterns" – design tricks that inadvertently guide users into making decisions they might not fully understand. Users can opt out of making all public prompts visible to only themselves or delete their prompts entirely via privacy settings, but many remain unaware of these options.

This situation highlights a critical gap between user expectation and the actual functionality of AI platforms, underscoring the ongoing challenge for tech companies to ensure full transparency and informed consent regarding user data and privacy in the rapidly evolving AI landscape.