ZK-JPEG: Zero-knowledge Image Editing and Compression
Cryptographers present ZK-JPEG, a zero-knowledge proof system verifying a published JPEG was correctly compressed from a signed camera image.
IACR ePrint paper 2026/2039 introduces ZK-JPEG, a cryptographic tool from researchers at Stealth Software Technologies and the University of Vermont. It uses zero-knowledge proofs to show a published image derives from a secret committed input while surviving lossy JPEG encoding, which invalidates camera attestation signatures. The system verifies a large family of image transformations, such as blurring and redaction, by integrating them into JPEG compression with minimal cost. It is built with PicoZK, which compiles Python image-editing code into circuits for the line-point zero-knowledge (LPZK) proof system, and is motivated by defending image provenance against deepfakes.
- Proves a published JPEG was correctly compressed from a secret committed input
- Survives lossy JPEG encoding, unlike prior zero-knowledge image-provenance schemes
- Verifies blurring, redaction, and other transformations within the compression process
- Built with PicoZK and the LPZK proof system from off-the-shelf ZK tools
- Motivated by camera attestation to counter deepfake photographs
- Organizations
- University of Vermont
From the original article
1 min read at eprint.iacr.orgTools for generating deep fake photographs are proliferating with greater ease of use and prominence in pop culture. Image authentication tools can defeat these deceitful developments by verifying that a digital image was actually produced by a physical camera. The challenge is that these tools must be robust to desirable image transformations.
By eprint.iacr.org via vaultdmn · eprint.iacr.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.