Researcher Releases Adversarial Patterns That Defeat Deployed Surveillance Cameras

At Def Con 2026 in Las Vegas, security researcher Bill Swearingen demonstrated adversarial patterns that, printed on vehicles, clothing, or objects, prevent commonly deployed surveillance cameras and license plate readers from detecting people, faces, and vehicles. The patterns scramble object identification without blocking video recording. Swearingen developed them over roughly 31 million tests across a year and released the designs at norecognition.org.