ChatGPT's background conversations auto-created an 'Abuse Confession' chat without user's knowledge
lisak40 · reddit · 2026-09-16
A Reddit user enabled ChatGPT's 'Start with Voice' and 'Background Conversations' options, never used them, and later found the app had autonomously created a standalone chat titled 'Abuse Confession' about a single instance of spanking a child years ago. The user worries it could trigger a report to authorities. The incident highlights unexpected autonomous behavior and privacy concerns in ChatGPT's background conversation feature.
More from Models
- LAB legal benchmark flaw: case docs leak planted issues directly to models — andersonbcdefg · 2026-09-17
- MiMo eval chart shows judge and probe disagree 60% of the time, sparking reward-hacking concerns — andrew_n_carr · 2026-09-17
- Fans mourn the end of Codex lead's regular daily reset cadence — kimmonismus · 2026-09-17
- Users report Google Astra feels noticeably degraded over past two days — pwlot · 2026-09-17
- OpenAI burns 20% as much compute on monitoring as the model itself, SemiAnalysis says — kevinnbass · 2026-09-17
- Google releases Gemma 3n: 2GB RAM multimodal model, first sub-10B to top 1300 on LMArena — joemeno · 2026-09-17