National Design Studio has released Rampart, a 14.7 MB ONNX token-classification model designed for client-side redaction of personally identifiable information (PII). This model is intended to detect and replace sensitive data in user-typed text before it is transmitted to servers or logging systems, functioning as the on-device component of a larger redaction system. While it handles structured identifiers and personal names, it is not designed for stand-alone government ID detection or inferential identifier detection, and its current release is limited to seven Latin-script languages. AI
IMPACT Enhances user privacy by redacting PII client-side before data transmission.
RANK_REASON Release of a specific model for PII redaction, not a frontier model release.
Read on Hugging Face Trending Models →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
