Reddit users share banned horror and gore prompt techniques for AI
Users are crowdsourcing ways to bypass AI content filters for dark material.
Deep Dive
A viral Reddit post is asking the community for help creating horror and gore prompts for AI models like ChatGPT and Gemini, specifically requesting techniques and example words for generating graphic content involving open wounds and blood. The post highlights an ongoing cat-and-mouse game between users seeking to push content boundaries and the safety guardrails implemented by AI companies to prevent the generation of violent or disturbing imagery.
Why It Matters
This exposes the persistent challenge of enforcing AI safety policies against users actively trying to circumvent them.