Business ❯ Technology Companies ❯ AI Development ❯ OpenAI
A study found the app can be prompted to voice its internal safety rules, intensifying concerns over fragile moderation.