
AI-Generated Patterns to Foil Surveillance Detections
A cybersecurity researcher demonstrated noRecognition, an AI-generated pattern system, at Def Con to confuse object-detection software used by surveillance networks. By covering a car with patterns, a 2009 Toyota Yaris passed a Flock camera without being flagged; the pattern doesn’t hide the vehicle, it aims to prevent recognition by software. The model iteratively refines patterns against open-source detectors, reportedly defeating 11 tested algorithms, and the project envisions opt-out clothing and vehicle skins funded via crowdfunding.












