Workaround for ChatGPT's stuck bio filter: delete the triggering past chats
davidmanheim · x · 2026-11-02
David Manheim follows up with a concrete fix for ChatGPT's stuck biological-risk filter: delete the past chats that triggered it, and probably also the new chats that produced the same refusals, to be safe. It's the most direct workaround for the glitch that blocks every prompt.
Related event: ChatGPT bio-safety filter blocks everything; deleting chat history fixes it(3 posts)→
More from Safety
- User gets safety warning after letting Antigravity scan his own servers for exploits — antiriad76 · 2026-11-02
- Alabama passes age-verification law 138-0 for all app downloads; Big Tech lobby sues — amplifiedamp · 2026-11-02
- Purser: an MCP server that lets agents pay with smart-contract limits a jailbreak can't bypass — Plane-Preference6245 · 2026-11-02
- "Copyright is for courts to decide, not for the tool to tell me what I can make" — adamamcbride · 2026-11-02
- US data companies quietly sell to Chinese AI labs — a shadow market with no oversight — karinanguyen · 2026-11-02
- Grok Confirms Chrome macOS Vulnerability Patched Since 155.0.8059.39 — emax · 2026-11-02