OpenAI adds an AI safety layer to detect misuse without retaining enterprise data

OpenAI is adding a new safety capability that allows enterprises to detect misuse of its AI systems across multiple interactions without retaining prompts or...