Bug Report: Flagged outputs now fully censored when conversation is resumed from history
`Steps to Reproduce`
Have a conversation with ChatGPT about a sensitive subject where it's likely to falsely throw a flag on part of the conversation without making it disappear. After getting a red flag, note that that part of the conversation is still visible, still there as part of the context and the conversation can be continued past that point in a coherent manner.
`Expected Result`
When loading the conversation from history, previously flagged outputs should still remain visible, along with the red warning.
`Actual Result`
Now, when an old conversation is reloaded, those flagged parts are missing completely, looking like ChatGPT had gone full censorship and removed the output immediately, when it had definitely not done so at the time of the flag.
`Environment`
Windows 11, Web version, using Firefox just recently updated to latest version yesterday.
fleet trenchBOT