1. Natural Input
Staff enter sensitive data through typing or dictation. Your audio or text is processed securely before anything is sent to the AI.
Give staff a practical way to use browser AI. Deidentified detects personal information and replaces it before prompts reach supported AI tools, so people can work with useful context, not identifiable details.
Use our browser extension to de-identify sensitive information when working with ChatGPT, Claude, Gemini, Perplexity, Grok and Copilot. You can also work in our separate web app, where your selected profile determines whether de-identification is switched on or off.
“Please draft a reply to Fiona Tan at fiona.tan@example.org about their application. ABN 51 824 901 773.”
“Please draft a reply to [PERSON] at [EMAIL] about their application. ABN [ABN].”
Removes sensitive information before content is sent to the AI, with staff reviewing the result.
“Please draft a reply to Fiona Tan at fiona.tan@example.org about their application. ABN 51 824 901 773.”
“Please draft a reply to [PERSON] at [EMAIL] about their application. ABN [ABN].”
Staff enter sensitive data through typing or dictation. Your audio or text is processed securely before anything is sent to the AI.
Deidentified extracts sensitive details and replaces them with intelligent placeholder tags - maintaining context for AI understanding.
Personal information is replaced as soon as it is pasted, so staff see exactly what the AI will receive, and it is checked again when the message is sent. The AI works only with placeholders to create clean, structured outputs.
When staff copy the AI's response out, Deidentified reinserts the original details, so the final document contains the real details.
With staff reviewing what goes in and what comes out, your organisation can use AI while reducing privacy risk and supporting its privacy obligations.
Watch how Deidentified protects your organisation's data in real time
Not ready to deploy a full tool yet? Start with the evidence. A 4-week assessment that shows how public AI is being used through the channels the Snapshot monitors and where sensitive data may be at risk.
Designed to support privacy practices under the Australian Privacy Act 1988 and New Zealand Privacy Act 2020, alongside data sovereignty requirements.

The Deidentified Chrome and Edge extension works directly in supported AI platforms, providing seamless de-identification protection where you need it most.
ChatGPT, Claude, Gemini, Perplexity, Grok and Copilot in your browser (not the Copilot desktop app)
Protect both free text and file uploads in real-time
Organisations can enforce automatic de-identification of pasted content before it enters the AI field

Organisations want the power of AI - the speed, the smarts, the efficiencies - but protecting customer privacy makes adoption very challenging.
Strict privacy regulations mean that in many organisations, public AI platforms are blocked entirely. This creates 'shadow AI' - the unauthorised use of AI in the workplace. Employees bypass controls to increase productivity, exposing organisations to serious privacy breaches and compliance violations.
In settings where AI platforms haven't been blocked, staff may be permitted to manually de-identify and re-identify information. It's more efficient than no AI use, but it's a slow, risky process - prone to human error and potential breaches.
Deidentified is a privacy-first gateway that solves the "block vs allow AI" problem by giving organisations a technical safeguard that lets staff use AI while reducing the risk of exposing personal information.
Use AI more safely while supporting your obligations under the Australian Privacy Principles (APPs) and the New Zealand Information Privacy Principles (IPPs).
Deidentified works seamlessly with any Large Language Model (LLM) platform through our secure API gateway or as an extension plug-in. Automatically strip sensitive information before it reaches any AI tool, ensuring your data remains protected while you leverage the power of modern AI capabilities.
Your personal information stays protected in Australia. Deidentified runs on AWS in Sydney, and on-premises deployment is available on request.
Provides a technical safeguard that can form part of the reasonable steps an organisation takes to protect personal information under APP 11 (Australian Privacy Act 1988) and IPP 5 (New Zealand Privacy Act 2020). Built for small or large organisations.
Tailored compliance for every Australian and NZ sector
Privacy-focused data protection for banks, credit unions, and fintech companies across Australia.
Secure AI adoption for Australian Federal, State, and Local government agencies.
Customer data protection and privacy compliance for large Australian businesses with >$3M annual turnover.
Student privacy protection for Australian and NZ universities, TAFEs and schools.
Donor and client privacy protection for Australian and NZ charities and not-for-profits.
Join Australian and NZ organisations using AI with privacy safeguards built in.
9:00 AM - 6:00 PM AEST
Designed to support APP 11 and IPP 5
Processing on AWS in Sydney
Expert assistance available
Get a personalised demo with enterprise-scale scenarios and compliance requirements