Remove PII before AI prompts
Detect names, email addresses, phone numbers, locations and other identifiable values, then replace them with stable placeholders that keep the surrounding sentence useful.
AI Text Anonymizer
Anonymize sensitive text before using it with AI.
anonymise.online is an online text anonymizer for AI prompts and LLM workflows. It replaces detected personal information, identifiers and secrets with context-aware placeholders before you paste the sanitized text into ChatGPT, Claude, Gemini or another AI system.
Protect API keys and credentials
Known secret formats and high-confidence token-like values are redacted before the text leaves your workflow.
Keep context for the LLM
Context-aware labels such as EMAIL, PERSON, IBAN or TOKEN preserve the role of a value without exposing the original plaintext.
Restore locally with your map
After AI processing, keep the anonymized text and private restore map separate. The fingerprint is reference metadata; restoration still works after AI edits when the original placeholders remain present.
Typical AI anonymization workflow
- Paste sensitive source text into anonymise.online.
- Review automatic English and Italian detection and add any manual redactions.
- Download the private restore map and copy only the anonymized text.
- Use the anonymized text in your AI assistant or LLM workflow.
- Restore the original values only when you need the final plaintext again.
No automatic detector is perfect. Review the output before sharing it with any third-party AI service.