Designers pitch "adversarial clothing" meant to confuse facial-recognition systems

As facial-recognition systems spread across public spaces in Britain, The Guardian reports on July 17, 2026 that a set of designers are selling “adversarial clothing”: garments printed with shapes, colours, and repeated motifs meant to exploit weaknesses in some computer-vision systems. The approaches described include large-scale prints, asymmetrical cuts, patterns that present a machine with many apparent faces at once, and, in at least one coat, infrared LEDs.

The brands named include Vollebak, whose co-founder Nick Tidball suggests the category may be near the mainstream, along with Cap_able and Urban Privacy. The article is careful about what the clothing does and does not do. Jennifer Bell, a senior lecturer at the Nottingham School of Art and Design, is quoted saying that “none of these products are tried and tested, and a lot of these surveillance technologies can deal with a little resistance.” Daniel Preuss of Urban Privacy grants that “no design can guarantee security from detection,” and frames the garments partly as a way to spread public awareness.

The Specola files the trend and its documented limits. The reported effect depends on the specific system and conditions, and none of the products has independent real-world validation. The interest here is not a working invisibility cloak but the appearance of a consumer market in resistance itself: the water-world beginning to dress against the eye it built.