Reimagining Moroccan Rugs Through AI

In COVER’s sister magazine HALI— concerning the subject of antique rugs and textiles—much of the text and discussion is focused on design evolution. In most cases the motifs, colours, knots and wool reveal a rug’s origins and also foreign influences on the region. Thanks to the Silk Road and other ancient trade routes, design ideas spread from area to area and weaver to weaver.

When Ida Hausner and Max Blazek began development for the MLRug (machine learning rug) project —a series of thirteen AI-designed Moroccan rugs—they were contemplating Artificial Intelligence’s relation to this ancient transference of design. Ida explains that they were asking the questions: ‘How does AI pattern development actually work? Could it perhaps be imagined similarly to the analogue migration of motifs in carpets?’

Ida and Max are the children of Grazbased Moroccan rug gallerist and expert Gebhart Blazek and, while rugs had been all around them in their childhoods, Ida describes them as ‘a major but until then passive presence in my life’. However, Max’s design theory course at the Weissensee Kunsthochschule Berlin, and Ida’s master’s thesis for her studies at the University of Applied Arts Vienna, brought them to a confluence of thought on the subject of AI.

In 2019, Max’s course looked at the use of artificial intelligence in design, ‘I found these theoretical considerations incredibly enriching, but I also wanted to tr y it out in practice,’ he says. Max asked his father for ‘at least 1,000 high-res images of Moroccan rugs, the more the better’, and began experimenting with algorithms. However, he found the resulting designs rather lacking, which he puts down to a shortfall of data, a common issue with AI inputting.

Cut to 2022, and Ida’s thesis looking at analogue and digital pattern development—with nods to inventor Charles Babbage and mathematician Ada Lovelace—inevitably turned her attention to Moroccan rugs. ‘What always fascinated me most was the abundance of colourful and seemingly chaotic patterns, derived from rug models from urban manufactories,’ she says.

Ida was keen to explore the ideas within her thesis further, and Max was happy to return to the project he had already begun. They both wanted to see AI in action, so they decided to create a rug collection to be exhibited. Ida says ‘I wanted to comprehend it with my whole body—as a handcrafted object that can be touched.’ With his knowledge of rugs and production in Morocco (plus a fantastic gallery to take over during May 2025’s Designmonat Graz) Gebhart joined the party.

Max returned to training algorithms, this time using his father’s high-quality images plus extra examples from the internet to help the process along: ‘It was important to us that these images were only used to train an algorithm within a purely scientific, artistic and transformative context,’ he explains. While in 2019 he used generative adversarial networks, or GANs, for generating designs, in 2022 he moved to StyleGAN2-ada, an advanced version but one that required less in terms of data/images compared to algorithms from Transformer architecture such as the often-discussed DALL·E 2 by OpenAI.

Production of the rug would have been much easier in India or Nepal, where manufacturers process designs digitally to make easy-to-use knotting templates; but the project stuck to its roots. ‘We opted for production in Morocco, with all its improvisational imprecisions,’ says Gebhart, ‘because this approach better reflects the character and vitality of the original material—and also ensures that the value creation remains in the country of origin of the design references.’

Exhibited in May at Gebhart’s Berber Arts gallery during Designmonat Graz 2025, the thirteen resulting rugs are replete with Moroccan rug motifs and have a wild energy that belongs to the tradition. This ancient craft has been reimagined digitally and then recrafted once again. The thread of the traditional rug story runs through each stage of the design and links into Ida’s thesis, in which she talks about how ‘ the digital world and textile craft already intersected at the beginning when programming itself took its first steps in the textile industry’.

Max and Ida recognise the potential pitfalls of working with AI, both personally and for the design world; as Max says, ‘The results are often not critically evaluated.’ Yet MLRugs is a project created from the curiosity of three individuals, trying to understand how craft and technology meet, and in so doing making a beautiful Moroccan rug collection. ‘With MLRug, our aim is not to copy, revolutionise or replace Moroccan rug culture,’ says Ida. ‘It’s about opening a new window and seeing where it might lead.’