OpenAI unveils “Privacy Filter,” a privacy protection model with a lightweight design for commercial use

OpenAI released Privacy Filter, an open-weight model for detecting and masking personal information in text.

The 1.5B-parameter model can run locally on laptops or in browsers and is available on Hugging Face and GitHub under Apache 2.0.

It is designed to help developers reduce privacy risks before sending data to AI systems.

The tool covers eight categories of personal data, but OpenAI says it still needs human oversight in sensitive cases.