Fashion meets AI: The future of an augmented industry The 9 October 2024 - Première Vision Paris - Blossom Première Vision - Première Vision New York

New machine learning models are reshaping the future of fashion, stirring emotions ranging from fascination to scepticism, enthusiasm to dismissal. This duality mirrors the divide between fashion and the fashion industry, rooted in their respective relationships with innovation and tradition: a pull to the future, technology and the digital world on one side, and the prestige of heritage, craftsmanship and physicality on the other. Instead of viewing them as opposing, mastering and merging these (seemingly) contrasting forces could pave the way to new approaches in designing, producing and managing the lifecycle of garments.

A recent study¹ by Bain & Company and the Comité Colbert highlights that luxury brands are currently exploring various generative AI applications. Key areas of focus include operational efficiency (such as sales forecasting, inventory optimisation, and quality control) and enhancing customer engagement (personalised marketing, vendor-client dialogue, and customer segmentation).

The global fashion market’s use of AI is projected to reach $4.4 billion by 2027².

Though AI-driven automation and computer visualisation are increasingly augmenting human intelligence and creativity, luxury brands have not fully embraced tools that upskill employees. Everything related to creativity remains delicate, and confidentiality, linked to security within processes is a crucial concern.

Read also: Can algorithms and artificial intelligence serve a more responsible fashion?

This three-part series of articles explores how AI is shaping the future of fashion in its ability to:

segment and analyse–more or less directed, and here lies the importance of human common sense and ethics–immense and unstructured amounts of data in record time (questions about the quality and veracity of this data, as well as confidentiality and security)
predict—more or less real—needs, trends (to follow or not), errors or defects, sometimes invisible to the human eye
generate content–which can be totally biased or innovative, unprecedented, unimaginable, even hallucinatory.

AI and the future of fashion — ©Google DeepMind

Reinvesting in the upstream chain

One of AI’s most promising applications in the fashion industry lies in improving traceability and transparency across the supply chain. By modelling workflows and enhancing communication between stakeholders (assuming interoperability and standardisation), AI makes it possible to guarantee the authenticity of fibres and textiles, while helping them meet non-financial reporting obligations and Extended Producer Responsibility (EPR) regulations.

Companies such as Fairly Made, Retraced AI, TrustTrace, CrystalChain and Carbonfact are already leveraging solutions for data security and traceability (using blockchain technology) throughout the value chain. AI identifies areas for improvement, mitigates risks, all while contributing to decarbonisation goals.

Read also: Smart Keys: How do digital tools accelerate traceability?

AI fashioning tomorrow — ©Google DeepMind

Above all, AI plays a pivotal role in minimising waste, both in textile and garment production, conserving resources and reducing wasted materials, time and money. Textile Genesis, a leader in traceability, integrates and maps over 300 “product flows” from fiber to finished product, each of which includes estimates of discarded materials generated along the way.

Back in the field, right down to the land, AI and satellite imagery are being used to monitor agricultural techniques and detect fraud. For example, GOTS is collaborating with the European Space Agency (ESA) and AI expert Marple to oversee organic cotton crops in India from space. Physical (forensic) traceability methods, such as those offered by Oritain, are valuable AI applications.

The sourcing, rendering and development of the materials themselves, whether virtual or real, will be the subject of our next article!

Revolutionising production

AI is also unlocking new possibilities in the design and production of clothing. Digital Product Creation (DPC) 3D tools allow designers to simulate multiple versions of a garment without physically producing each prototype. This process not only speeds up the creative design process, but also reduces overproduction and the waste resulting from unused models.

By integrating AI into smart production machines, companies like Lectra, with its unique portfolio of technological and SaaS solutions for the textile industry, optimise fabric cutting by adjusting pattern layouts in real-time, reducing material waste by up to 30%. This improved management of raw materials is critical in terms of minimizing the textile industry’s environmental impact.

Generative AI is able to create and simulate an infinite number of variations in record time, drawing from a vast, preferably intelligently curated database. It thus expands the range of what is (more or less) possible, far beyond what humans could achieve. However, human creativity remains essential—knowing not only how to create the right “prompts” and merge traditional and Web 4.0 techniques, but above all how to discern which elements to embrace and which to discard amidst the endless possibilities.

One noteworthy example is imki, a company specialised in developing bespoke AI solutions tailored to a brand’s DNA. In collaboration with Kipas Textiles and Taypa, world leaders in denim manufacturing, imki conceptualised, designed, and produced a denim capsule collection in just 21 days, featuring three silhouettes and six contemporary, original pieces.

Ultimately, the physical realm remains crucial. Generative AI such as Wearfits or Truefit tackle the highly complex task of digitally rendering the human body—its physical reality, composed of multiple supple, malleable and sensitive layers, with as many unique characteristics as the eight billion human beings on the planet.

Read also: Can avatars assist a sustainable approach?

AI’s ability to analyse behavioural data, such as buying preferences and social media trends, helps predict sales in real-time and adjust production ever more closely to the “needs” of an increasingly impatient consumer. As Sage Lenier, the executive director of the climate nonprofit Sustainable and Just Future, puts it, “AI enables fast fashion to become the ultra-fast fashion industry”. It is quite literally, efficiency on steroids. But the environmental cost is staggering.³

The environment, humanity and quality, along with the beauty and meaning of fashion itself, are the true sacrifices of this model. This “on-demand” production model bears little resemblance to bespoke or haute couture, which represents slowness, traditional craftsmanship, artisanal practices, genuine personalisation and sustainability.

In our third AI-focused article, we delve deeper into how these emerging technologies can positively impact garment manufacturing at the dawn of a fourth industrial revolution.

Reinventing the downstream chain

The end-of-life management of garments is another area where AI plays a crucial role in fostering a transition towards a truly circular economy. Currently, most clothing ends up in landfill due to its growingly poor quality and the complexities of recycling blended materials. AI helps overcome these challenges by facilitating the identification, sorting and recycling of textiles. The CETIA recycling centre, inaugurated in 2023, brings together automated sorting technologies and end-of-life dismantling in a single facility.

Resortecs, a company specialising in textile recycling, has developed a heat-soluble thread that allows garments to be easily disassembled. AI is used to optimise the dissociation temperatures and automates the sorting of components for recycling.

While promising, these systems remain costly and require sophisticated infrastructure. Implementing an AI-driven circular economy involves significant investment in technology and infrastructure, as well as collaboration between brands, consumers, and governments to scale these solutions effectively.

Read also: Smart Key: Can the rapid acceleration of fashion-tech help achieve eco-responsible goals?

While digitising workflows in fashion is essential, we must still question whether AI can save the planet while feeding the demand for the latest trends, given its energy-intensive nature⁴ and carbon footprint⁵.

However, while the issue of managing its environmental impact and the ethical and legal risks it entails remain central concerns, artificial intelligence is nonetheless a tool with immense potential. Nevertheless, if AI is implemented within a well-structured, sustainable, long term strategy—addressing the real needs of all stakeholders, from farmers and biotechnicians to spinners, weavers, manufacturers, and consumers—it holds immense potential.

“The key question is how to develop technologies that meet our human needs without destroying the ecosystems that enable other creatures to meet theirs.”
Tarik Chekchak,
Director of the Biomimicry Unit at the Institute of Desirable Futures