OpenAI wants to monitor AI abuse without forcing customers to hand over their data

OpenAI’s Private Safety Processing promises to detect abuse across multiple interactions while preserving Zero Data Retention, putting its privacy strategy in direct contrast with Anthropic’s 30-day retention requirement.