News
Berlin Artist Designs Shirt That Fools AI Surveillance Cameras
Simon Weckert created a patterned Hawaiian shirt that, in tests, confuses the YOLO algorithm and makes it harder for AI systems to detect a human figure. The project was made in response to a pilot AI surveillance system launched in August 2026 at Berlin's Kottbusser Tor square.
Contents
Berlin digital artist Simon Weckert has designed a shirt called Digital Camouflage, whose intense, chaotic pattern is meant to stop AI-equipped surveillance cameras from recognizing the wearer as a human being. The project is a direct response to a pilot AI-based video surveillance system that Berlin authorities launched in August 2026 at Kottbusser Tor square in the Kreuzberg district.
How the camouflage works
The shirt features a chaotic pattern in saturated shades of pink, orange and green, arranged in overlapping shapes. Instead of hiding the body under a solid color, the pattern deliberately overloads the early stages of image analysis, where the algorithm tries to extract edges, contrast and the outline of a figure. As a result, the system can't stitch the individual fragments of the image into a coherent human form.
The saturated color transitions strongly activate the early layers of the network, and the overlapping shapes break the continuity of the body's contour, so the detector can no longer combine the individual parts into a single figure - Simon Weckert, creator of the Digital Camouflage project
Weckert developed the patterns through trial and error, using YOLO, a popular open-source model for real-time object detection that's also widely used in commercial surveillance systems, to test them. Successive versions of the pattern were checked against the same detector until the number of correct human-figure detections dropped clearly below the norm.
Testing at Kottbusser Tor
The choice of testing location carries symbolic weight. Kottbusser Tor is one of Berlin's most recognizable, and also most controversial, squares, long associated with debates over public safety in the Kreuzberg district. In August 2026, city authorities launched a pilot AI-driven video surveillance system there, which drew criticism from some residents worried about the expansion of surveillance in public space.
Weckert tested his design on a camera at that very location even before the system was formally switched on, treating the test as a form of artistic protest against the planned surveillance. According to the artist, the shirt is meant to be more than just a technical gadget.
AI systems never learned what a human being is, they learned what a human being statistically looks like across millions of training images - Simon Weckert
A growing anti-surveillance fashion market
Digital Camouflage isn't an isolated case. As facial-recognition cameras spread through shops, offices and public spaces, more designers are proposing clothing meant to confuse recognition algorithms. Italian brand Cap_able sells a reversible dress, one side of which is designed to make the recognition system identify the wearer as a zebra, for about $780. British brand Vollebak offers a radio-wave-blocking jacket for $4,295 and a windbreaker made of copper fabric for $1,895.
Makers of this kind of clothing caution, however, that none of these solutions works universally against every camera and AI model at once, and the patterns themselves quickly lose effectiveness as recognition systems are retrained on new data. Effectiveness also depends on camera quality, lighting and the specific algorithm installed in a given system.
The limits of effectiveness
Weckert himself treats the project more as a manifesto than a certified privacy tool. He stresses that the shirt doesn't guarantee anonymity under all conditions, and that its main goal is to draw attention to how easily video surveillance infrastructure is being expanded in cities today without much public debate. The AI-powered video surveillance market is already valued at more than $4 billion, with forecasts putting it at $10 billion by 2032, which is further fueling interest in this kind of artistic and commercial project.
For Polish readers, the topic isn't just an exotic curiosity from abroad. AI-assisted urban surveillance systems are also being tested and deployed in Poland, and the growing number of image-analysis-enabled cameras in cities, housing estates and shops means questions about the limits of surveillance, and how to curb it, are likely to come up more often going forward.
