Files you check
- Images are sent to our server and analyzed by detectors running on that server. When extra backup detectors are used, the image may also be sent to the Hugging Face Inference API (see Hugging Face’s own privacy policy).
- Videos: your browser extracts six still frames and a small slice of the file’s metadata for the check. After the result appears, the full video is uploaded in the background and kept with your other uploads for at most 24 hours.
- Vision AI review: to catch AI videos that pattern detectors miss, frames (and sometimes images) may be sent to Google’s Gemini API for a second opinion. Google may use data sent through its free API tier to improve its products; see Google’s Gemini API terms.
- Text is sent to our server and analyzed the same way.
- Temporary copies: the images, video frames and text you check are kept for at most 24 hours so the site owner can review wrong results and prevent abuse, then deleted automatically. They are never published or used to train models.
- We keep an anonymous fingerprint (hash) of checked images and the detector scores for 45 days, so the same image doesn’t need to be re-analyzed. The fingerprint cannot be turned back into the image.
Shared results
If you tap Share, we save the result and a small, low-resolution preview (or a short text excerpt) so the link works. Shared links may appear on our Recent scans page. Don’t share images of private people without their consent.
What we don’t do
- No accounts, no tracking cookies, no selling of data.
- No use of your files to train models.
- No storing of your files beyond 24 hours (unless you tap Share).
Contact messages & the metadata cleaner
- Messages sent through the contact form (your name, email and message) are kept until we have dealt with them, and are only used to reply to you.
- The photo metadata cleaner runs entirely in your browser — those photos are never uploaded.
Basic logs
Like every website, our server keeps short-lived technical logs (such as IP address and time) to prevent abuse. We count scans to show daily totals, and count visits together with the website that linked to us (for example Google or Reddit) and the page you arrived on — without cookies and without storing your IP address.