Privacy Policy

In one line. Pictures you pick from your device are scored inside your own browser. They are never sent to us. The plagiarism checker is different: it extracts a document's text locally, then sends passage queries through our server to a web-search provider. No advertising cookies or ad networks run here.

1. The pictures you check

When you drop in or select a file, it is decoded and scored on your own hardware by a model that was downloaded to your browser. The picture data never crosses the network and never reaches a server of ours. Reload or close the tab and it is gone from memory. We hold no copy, and no mechanism exists by which we could retrieve one.

This is worth stating precisely, because it is stronger than a promise. There is no upload endpoint for pictures on this site. You can confirm that yourself: open your browser's network panel, run a check, and watch that no request carries your file.

2. The one exception — pasting a link

Paste a web address rather than choosing a file and your browser first tries to fetch it directly. Many sites block that with a cross-origin restriction, and when they do we retrieve the resource through a third-party public proxy so the check can still run. In that situation the address you pasted is sent to that proxy so it can fetch the content. A file from your own device is never involved. If you would rather no third party sees the link at all, save the picture first and check the file instead.

3. Text used by the plagiarism checker

A PDF, Word or text file chosen in the plagiarism checker is opened in your browser and converted to readable text there. The binary file itself is not uploaded. When you press “Check for plagiarism”, the extracted text is sent over HTTPS to our Cloudflare Pages Function, divided into exact passage queries, and forwarded to the configured web-search provider. This transmission is necessary to search public online sources.

We do not add that text to a submissions archive or a training collection, and the endpoint tells browsers and shared caches not to store its response. Cloudflare may process ordinary request metadata and the search provider receives the query passages under its own terms and privacy policy. Do not use the checker for confidential, embargoed or restricted writing unless those handling rules permit this processing.

4. What we actually receive

• Messages you send us. Using the contact form passes the name, email address and message you typed to our form provider (Web3Forms), which forwards it to us by email so we can reply. We use it for that and nothing else.
• Ordinary request logs. Our host (Cloudflare) processes standard request data — IP address, user agent and the like — in order to serve the site and resist abuse, as every website does.
• Plagiarism-search text. Our server endpoint processes the extracted document text long enough to issue passage queries and return source results. We do not provide a database that stores it.

5. Cookies, storage and measurement

No advertising cookies are set, no ad scripts run, and no third-party analytics is loaded on this site at present — the tag is switched off in configuration, so the request is not made at all rather than being made and ignored. If that changes, this section changes with it; it is generated from the same setting the page itself reads, so the two cannot drift apart.

The tool does keep a few small, non-identifying things in your browser's own local storage, on your device: your light or dark preference, the running count behind the daily allowance, and result summaries you have saved. Media results may include a thumbnail. A saved plagiarism result contains its label, score, band and source count, not the submitted document or search snippets. Clearing site data removes those records.

6. Third parties involved

• Cloudflare — hosting and delivery for the site and for the detector's model files.
• Google Fonts — the typefaces are loaded from Google's font CDN, which may log the request, including IP address, in order to serve the file.
• Web3Forms — handles contact-form submissions on our behalf.
• A public fetch proxy — used only for the link case described in section 2.
• The configured web-search provider — Brave Search, or Google Programmable Search for an existing configured account, receives exact text passages only when you run the plagiarism checker.

Each of these operates under its own privacy policy. We do not sell, rent or trade personal data to anybody, and there is no advertising relationship behind this site.

7. How long anything is kept

Contact emails are kept only as long as your enquiry and any sensible follow-up require. There is no database of checked pictures because none of them ever arrive here. We do not operate a submissions archive for plagiarism checks. Third-party infrastructure and search-provider retention are governed by their policies.

8. Your rights

Depending on where you live you may have rights to access, correct or delete the personal data we hold about you — which, in practice, means the contents of any message you have sent us. Email hello@originalorai.com and we will sort it out.

9. Children

This service is not directed at children under 13, and we do not knowingly collect personal data from them.

10. Changes to this policy

This page may be updated; the date at the top always reflects the version you are reading. Anything material will be spelled out here rather than slipped in.

11. Getting in touch

Questions about any of this? Email hello@originalorai.com or use the contact page.