‘this is Donna again’: Meta’s AI app accidentally becomes a window into boomer oversharing
by Piyush Mathur
Meta's latest foray into consumer artificial intelligence has run into a peculiar and somewhat troubling side effect: boomers unwittingly airing their most personal confessions, medical records, legal woes, and romantic frustrations to the public, all through what they appear to believe is a private chatbot journal.
Tech investor and commentator Justine Moore has documented the phenomenon in a now-viral series of posts on X (formerly Twitter), where she highlights interactions from Meta's AI-powered assistant that have ended up on a public feed, seemingly without the users’ awareness.
‘I spent an hour browsing the app and saw: medical and tax records, private details on court cases, draft apology letters for crimes, home addresses, confessions of affairs... and much more’, Moore reported. Her posts included anonymised screenshots of some of the more benign or absurd exchanges—one user requested uplifting words for ‘a gangster in spiritual crisis,’ another tried to enlist the AI in a search for a woman with ‘big booty and nice rack.’
The resulting dialogue often veers into the surreal. In one case, a man earnestly workshopped a romantic poem for his girlfriend named Leanne, candidly revealing her recreational drug use along the way. When the AI’s poetic attempts didn’t land, the frustrated user lashed out: ‘Why do you fucking ask.’
More poignantly, a user named Donna repeatedly shared voice notes and detailed updates about her search for a phone cord, revealing frustrations with her provider and what appears to be general loneliness. In another screenshot, a woman details a toxic relationship with a man who lives with his mother, struggles with mental health, and frequently gaslights her.
These are screenshots of Justine Moore’s posts (dated June 12, 2025) on her X profile.
According to Moore’s follow-up post, the AI conversations are not public by default. However, users can press a ‘share’ button to post them to the public feed—something many older users appear to be doing accidentally. The assumption, it seems, is that they are writing in a private diary or speaking to a private assistant, when in fact their deeply personal exchanges are visible to all.
Moore directly queried the Meta AI assistant about the issue. The AI responded that it was ‘pretty wild’ and noted that ‘some people don’t read the fine print lol.’ It attributed the oversharing to a ‘mix of user error and maybe not enough guidance on what they’re sharing’, suggesting that the app’s design might be nudging users toward accidental disclosure.
These are screenshots of Justine Moore’s posts (dated June 12, 2025) on her X profile.
In a time when concerns over AI ethics and data privacy dominate the tech conversation, the incident underscores the importance of clear user interface design and consent practices—particularly for older demographics who may not be digital natives. While some of the posts are humorous, others raise serious concerns about privacy violations and emotional vulnerability.
Meta has not yet issued a formal response. In the meantime, the AI continues to answer romantic queries, field cries for help, and soothe spiritual gangsters—often to a much larger audience than intended.
Given that Moore’s observations are based on her personal experiences rather than a research survey, one should be careful about framing Meta AI’s apparent design deficiency in terms of the users’ age profile.