Digital Camouflage: The Shirt Designed to Make AI Surveillance “Blind”
This Hawaiian Shirt Tricks AI Surveillance Cameras Into Thinking You’re Not Human
In an era when AI-powered security cameras in cities can track locations and identify people, a designer has developed an unusual way to confuse computer-vision systems. Called “Digital Camouflage,” the shirt features deliberately strange, highly saturated colors and chaotic patterns. When a person wearing it passes in front of an AI camera, the visual pattern can interfere with the algorithms’ ability to recognize the wearer as a human, causing the system to misclassify or fail to detect the person. Although the shirt looks bizarre and unattractive to the human eye, it demonstrates how adversarial patterns can exploit weaknesses in AI-based surveillance and object-detection systems. The project is intended primarily as a commentary on automated surveillance and the reliability and transparency of AI systems, rather than as a guaranteed method of avoiding surveillance: https://petapixel.com/2026/08/27/this-hawaiian-shirt-tricks-ai-surveillance-cameras-into-thinking-youre-not-human/
Info Source:
The project is Digital Camouflage, created by Berlin-based digital artist and technologist Simon Weckert. His official project page explains that the garment uses an adversarial attack—a deliberately engineered visual pattern intended to manipulate what computer-vision models perceive. Demonstrations use the open-source YOLO object-detection system, where the wearer can fail to be classified as a person. Importantly, Weckert says the project does not guarantee anonymity or claim to defeat specific government surveillance systems: https://www.simonweckert.com/digitalcamouflage.html
Video: https://www.youtube.com/watch?v=A5QTnUlgcJQ
YOLO Open Source: https://github.com/ultralytics/ultralytics/blob/main/ultralytics/models/yolo/__init__.py