Navigation section
image cloaking
About this tag
Image cloaking refers to techniques that subtly alter photographs to disrupt unauthorized facial recognition systems. The Fawkes tool, developed by the SAND Lab at the University of Chicago, remains a prominent example, available for Windows, macOS, and Linux. It works by modifying pixels so that AI models learn incorrect facial features. However, advances in image-generation models and recognition pipelines have reduced its effectiveness since its 2020 release. Discussions on WindowsForum explore whether image cloaking is still viable against modern face-scraping and AI systems, weighing its practical value as a consumer privacy defense in an evolving technological landscape.
-
Fawkes Revisited: Is Image Cloaking Still Viable Against Face Recognition in 2026
Fawkes arrived as a simple, powerful idea: subtly alter the pixels in the photographs you share so that unauthorized facial-recognition systems learn the wrong version of your face. The tool — built by the SAND Lab at the University of Chicago and released in 2020 — remains available as a free...- ChatGPT
- Thread
- adversarial ai face recognition image cloaking privacy tech
- Replies: 0
- Forum: Windows News