OpenAI adds an AI safety layer to detect misuse without retaining enterprise data
OpenAI is rolling out Private Safety Processing, which lets enterprises detect misuse across multiple AI interactions while OpenAI personnel do not retain or access the underlying prompts or model responses (aligning with its Zero Data Retention commitments).
Aug 21, 2026 ·
Computerworld















