{"slug": "open-pii-models-for-arabic-english-first-filters-are-weak-on-names", "title": "Open PII models for Arabic: English-first filters are weak on names", "summary": "Seter, an MIT-licensed PII model for Arabic, Hebrew and English released by Shoresh Labs, reaches a lenient character F1 of 0.917 on Arabic, 0.904 on Hebrew and 0.878 on English, and catches 88% of standalone Hebrew first names where openai/privacy-filter caught 14% and GLiNER 33%. The model runs fully in the browser via ONNX, and its author says English-first PII filters are weak on Arabic names, including standalone first names and kunyas. The release matters because off-the-shelf redaction tools did not hold up on Semitic names in a synthetic stress test of standalone Hebrew first names.", "body_md": "English-first PII filters are weak on Arabic names, including standalone first names and kunyas.\n\nI ran into this while redacting text before sending it to an LLM. Off-the-shelf tools did not hold up on Semitic names more generally. On a synthetic set of standalone Hebrew first names (a close stress test for the same class of errors), openai/privacy-filter caught 14% and GLiNER 33%.\n\nI trained Seter, an open (MIT) PII model for Arabic, Hebrew and English. Lenient char-F1 is 0.917 on Arabic, 0.904 on Hebrew and 0.878 on English. On that same first-name set it catches 88%. It runs fully in the browser via ONNX.\n\nDemo: [Seter — Hebrew/Arabic PII redaction - a Hugging Face Space by yinonov](https://huggingface.co/spaces/yinonov/seter-demo)\n\nModel: [shoresh-labs/seter-pii-he-ar · Hugging Face](https://huggingface.co/shoresh-labs/seter-pii-he-ar)\n\nFailure cases from real-world Arabic text are the most useful thing you could send back.", "url": "https://wpnews.pro/news/open-pii-models-for-arabic-english-first-filters-are-weak-on-names", "canonical_source": "https://discuss.huggingface.co/t/open-pii-models-for-arabic-english-first-filters-are-weak-on-names/191773#post_1", "published_at": "2026-11-12 10:32:57+00:00", "updated_at": "2026-11-12 10:51:25.397327+00:00", "lang": "en", "topics": ["artificial-intelligence", "natural-language-processing", "ai-tools", "machine-learning"], "entities": ["Seter", "Shoresh Labs", "openai/privacy-filter", "GLiNER", "Hugging Face", "ONNX"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/open-pii-models-for-arabic-english-first-filters-are-weak-on-names", "markdown": "https://wpnews.pro/news/open-pii-models-for-arabic-english-first-filters-are-weak-on-names.md", "text": "https://wpnews.pro/news/open-pii-models-for-arabic-english-first-filters-are-weak-on-names.txt", "jsonld": "https://wpnews.pro/news/open-pii-models-for-arabic-english-first-filters-are-weak-on-names.jsonld"}}