Sora user stuck for months as OpenAI export page throws HTTP 500 and original Library is gone
Realistic_Winner_493 · reddit · 2026-11-20
A former Sora user describes months of frustration with OpenAI's data export after the shutdown: official exports were inconsistent, the original Sora Library is gone so there's no way to verify what's missing, and his manual backup was lost to a hard drive failure.
The export site now fails entirely for him, with auth requests repeatedly returning HTTP 500. Support has endlessly escalated, deduplicated, and routed his case to the Privacy team without answers. He asks why OpenAI couldn't keep a read-only Library or publish a list of historical creations, and pleads for one last reliable way to recover his own work.
Related event: Sora users stranded as export tool fails months after shutdown(2 posts)→
More from Companies & People
- Microsoft Research podcast: Eric Topol, Eric Horvitz on AI reshaping biology and medicine — tw_killian · 2026-11-20
- Johns Hopkins hosts free webinar on improving academic research productivity with AI tools — JeremyNguyenPhD · 2026-11-20
- Bridgewater's Greg Jensen backed Anthropic and OpenAI early, but sold his OpenAI stake for succession plan — nordicinst · 2026-11-20
- Gary Marcus: that three-week-old tweet may become the worst-aging tweet in X history — GaryMarcus · 2026-11-20
- AI commentator slams Claude's EA-affiliated team: "You can actually feel the evil" — AIFlow_ML · 2026-11-20
- Google Now Ships Two Models a Week, Six of Top 10 Spots on Artificial Analysis Index — haider1 · 2026-11-20