According to Bloomberg, a Reddit user posted a photo (the image) displaying brief descriptive titles of several ChatGPT conversations that did not belong to them. The same bug was also captured in a screenshot shared by another individual on Twitter. OpenAI's spokesperson confirmed the incident, clarifying that only brief descriptive titles, not full transcripts of conversations, were exposed due to the bug.