Alexander Nemecek, a PhD candidate in the Department of Computer and Data Sciences at Case Western Reserve University, received the best paper award at the Authenticity and Provenance in the Age of Generative AI (APAI) Workshop at the IEEE / CVF Computer Vision and Pattern Recognition (CVPR) Conference 2026.
Held earlier this year in Denver, Colorado, this conference is one of the world's premier events for computer vision and artificial intelligence research. CVPR also is among the most selective and highly cited venues in computer science.
Nemecek’s paper titled, “Authenticated contradictions from desynchronized provenance and watermarking,” was co-authored by:
- Erman Ayday, PhD, associate professor of computer and data sciences at Case School of Engineering;
- Guang Cheng, professor of statistics and data science, computer science and biostatistics at the University of California, Los Angeles; and
- Hengzhi He, a PhD student in statistics and data science at the University of California, Los Angeles.
As part of the award, Nemecek received a $1,000 prize sponsored by Pangram, a startup specializing in AI-generated content detection.
About the paper
“Authenticated contradictions from desynchronized provenance and watermarking” identifies a vulnerability the authors call the “integrity clash.” The vulnerability occurs when an AI-generated image carries a cryptographically valid content credential claiming human authorship while its pixels simultaneously contain an invisible watermark identifying it as AI-generated—allowing both verification checks to pass independently.
The team demonstrated that these "authenticated fakes" can be created using standard editing tools, then proposed an audit protocol that detects the contradictions with 100% accuracy. As provenance labeling and AI watermarking see wider adoption across platforms like Adobe, Google and Meta, the work highlights a gap that content authenticity systems will need to close.
