Privacy-first workflow
SanitiseAI processes sensitive text in a focused workflow designed to minimise retention risk. Content is sanitised during request handling and returned as safe-to-share structured output.
Browser extension
One focused action, with clear boundaries around processing and storage.
The text you choose is sent over HTTPS when you click Sanitise, use the selected-text action, or enable automatic sanitisation for a supported form submission.
Text may include personal details or credentials. Processing happens on the server, not on your device.
The extension stores its automatic-mode preference in browser storage. It does not keep a prompt history or include analytics or advertising scripts.
The info control shows detected categories and counts so you can understand what changed.
Core technology
Pattern heuristics and entity-based detection work together to reduce sensitive value leakage.
Entity detection improves identification in unstructured notes, contracts, and support transcripts.
Post-processing checks reduce duplicate placeholders and obvious formatting leaks in anonymised output.
With your permission, we load Google Analytics on public information pages to measure visits. It is not loaded on the sanitiser or contact form, or on URLs containing query strings or fragments. Our page-view events exclude uploaded files, source text, sanitised results and contact messages.
Google receives technical connection information when its tag loads and may set analytics cookies. Advertising storage, advertising personalisation and Google signals are disabled in our tag configuration. We remember your choice in this browser for up to 180 days. Rejecting analytics does not affect the tool. Change or withdraw your choice using Analytics preferences in the footer; withdrawing also reloads the page to unload the tag.
From ingestion to cleanup, privacy controls are applied across the full request lifecycle.
Input arrives through encrypted transport and is handled in active request memory.
Detection and anonymisation run in transient request scope without saving the raw document.
Structured placeholders are returned so content remains useful while sensitive data stays protected.
Temporary buffers are released after request completion as part of the normal processing lifecycle.
SanitiseAI supports safer handling of sensitive text through practical controls, documented behaviour, and reviewable operational safeguards.
Ask what contact or support records we hold about you.
Ask us to correct contact details submitted through support or enquiry forms.
Request deletion of contact or support records where operational and legal retention allows.
Request export of contact or support records in standard formats when applicable.
Run SanitiseAI on real content, then contact us for rollout and policy guidance.
