ChatGPT keeps rejecting innocent game-art prompts as 'harassment and bullying'
AnonymousForALittle · reddit · 2026-10-27
A Redditor complains that ChatGPT repeatedly refused to generate archery longbow pose reference images for their game, citing guardrails around "harassment, discrimination, bullying" (screenshot included). Another case of image-generation safety guardrails misfiring on benign creative requests, highlighting how crude context judgment still is.
More from Fun
- GiveDirectly runs ad mocking Bryan Johnson to steer his billions to poverty relief — xuanalogue · 2026-10-27
- YC poker night: founders hit the tables, KlausCodes jokes he lost his chips to Jason Wang — KlausCodes · 2026-10-27
- "Our models can solve Millennium Problems, and I use them to archive spam" — craigsdennis · 2026-10-27
- Google escalates SERP poisoning: 40% of AccuRanker fetches hit at peak — SEO · 2026-10-27
- Demis Hassabis' "this is embarrassing" tweet resurfaces amid AI security chatter — egrefen · 2026-10-27
- The 'GPU-poor' hack: befriend top-university professors and casually mention you lack compute — vxnuaj · 2026-10-27