AI text anonymiser
Detect names, emails, IDs, secrets, addresses, and other sensitive details before text reaches AI tools, documents, or external workflows.
Structured redaction
{ "user": "John Doe", "email": "john.doe@example.com", "ssn": "999-01-2234" }{ "user": "[Person 1]", "email": "[Email 1]", "ssn": "[ID 1]" }The SanitiseAI standard
The tool replaces sensitive values with consistent placeholders, so the text remains useful after sanitisation.
Open the tool, load an example, or paste your own content without going through a setup flow first.
Text is sent to the sanitisation API over HTTPS for processing. Raw input is not stored by the tool.
Privacy by designChoose automatic mode for broad coverage or custom mode when you need precise control over entity types.
Choose detection rulesCommon questions
SanitiseAI is a text anonymiser for replacing sensitive information with readable placeholders before content is shared with AI tools, documents, or external workflows.
The sanitiser can detect common personal and operational identifiers including names, emails, phone numbers, addresses, dates, IDs, invoices, usernames, IP addresses, and secret keys.
Teams use SanitiseAI for legal drafts, support tickets, medical notes, finance summaries, developer logs, contract excerpts, and prompt preparation.
The output keeps the original structure where possible, replacing sensitive values with tokens such as [Person 1], [Email 1], or [Secret 1] so the text remains useful.
High-signal use cases
Prepare prompts for ChatGPT, Claude, Gemini, and internal assistants by replacing names, emails, addresses, and account details with readable placeholders.
Clean contract excerpts, renewal clauses, redlines, invoices, and matter notes before external review while preserving commercial context.
Remove API keys, passwords, tokens, usernames, IP addresses, hostnames, and stack-trace identifiers before debugging with AI or vendors.
Anonymise referral notes, support transcripts, customer escalations, HR cases, and operational summaries before sharing outside the original context.
Paste content, run detection, and copy structured output in one focused workflow.
Structured placeholders • Focused browser workflow • Share-ready output