{
  "id": 4871181,
  "title": "This 'Digital Camouflage' Shirt Confuses AI-Powered Surveillance Cameras",
  "url": "https://urgent.news/2026/09/01/this-digital-camouflage-shirt-confuses-ai-powered-surveillance-cameras",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-01T13:27:26.000Z",
  "source": {
    "name": "404 Media",
    "slug": "404-media",
    "url": "https://www.404media.co/this-digital-camouflage-shirt-confuses-ai-powered-surveillance-cameras/"
  },
  "original_language": "en",
  "account": "An artist named Simon Weckert has created a digital camouflage shirt designed to trick AI-powered surveillance cameras. When he wore the shirt in front of a camera, the AI failed to detect a person, revealing the shirt's effectiveness. Weckert developed the shirt as a response to the increasing use of AI surveillance cameras that can detect various objects and behaviors, including people. These cameras, already common worldwide, have recently been deployed in Berlin, raising concerns about privacy and potential discriminatory targeting of vulnerable groups like the homeless. Weckert's shirt utilizes a variation of YOLO, an open-source image recognition algorithm, and employs techniques to confuse the AI. By altering patterns, colors, and orientations, the shirt successfully thwarts the algorithm's ability to recognize a person. However, the effectiveness of the shirt depends on the specific AI algorithm used by the surveillance cameras, which is often unclear. Weckert's project not only aims to help individuals avoid detection but also to educate people about AI surveillance and provide alternative means of protest.",
  "summary": "I watched Simon Weckert's 'digital camouflage' in action.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}