Skip to content

Forensic Watermarking

BrightBlur embeds an invisible watermark into every photo you view. If a screenshot is leaked, the source can be traced.

When your browser decrypts a photo, it injects a steganographic watermark containing your unique identity into the rendered image. The watermark is imperceptible to the human eye but can be extracted computationally.

Because watermarking happens in your browser at display time, the server never sees the unblurred, watermarked image — maintaining the zero-trust model.

The watermark can be extracted from a leaked screenshot. The extraction is blind — it does not require the original photo. It can survive moderate cropping, compression, JPEG re-encoding, and resizing.

Watermarking provides accountability, not prevention. A determined leaker can photograph the screen with another device or apply heavy image manipulation to strip the watermark. But for the common case — a casual screenshot shared in a WhatsApp group — it provides a clear audit trail.

The goal is to ensure that everyone granted access to a private face understands that their viewing is traceable, which changes behaviour even when prevention is impossible.