This essay is part of the series: World Creativity and Innovation Day 2026: Sparks and Shields
In a 2025 interview, Jorgen Andersson, H&M’s chief creative officer, said: “We’re exploring emerging technologies like generative AI to amplify creativity and reimagine how we showcase fashion.”[1] Andersson was not alone. The previous year, 73 percent of fashion executives worldwide had said that generative AI—and AI more broadly—would be a priority for their businesses.
The growing focus on AI in fashion has translated into massive investments and experiments with AI integration across the value chain. The global AI in fashion market is presently worth around US$ 2.47 billion, and is expected to grow to US$ 9.45 billion by 2030 at a compound annual growth rate of 40.8 percent.
The change in corporate mindsets is palpable, too. In the last two years, AI has gone from being a priority to a necessity. As McKinsey’s State of Fashion 2026 report finds, executives now cite AI as the “biggest opportunity for the industry”, surpassing discrete aims such as product differentiation and strengthening sustainability. Indeed, AI is transforming not just design, manufacturing, marketing, and other functions, but the very workflows that could help the fashion industry minimise waste, and become more lean, agile, responsible, and resource-efficient.[2]
More than in Vogue
The drive towards sustainable fashion is a major imperative today. For fashion businesses, this stems from both a supply-side push to boost efficiencies and build an image of a responsible corporate citizen guided by the triple-bottom-line approach, as well as a demand-side pull where customers actively want more eco-friendly and ethically made apparel. Against this backdrop, AI is proving invaluable for embedding sustainable practices across the three stages of fashion operations: pre-production, production, and post-production.
Pre-production
Pre-production refers to the processes undertaken before a garment is mass-manufactured, such as design, pattern-making, sampling, and trend forecasting. The use of AI to support some of these functions has opened new avenues for promoting sustainability.
AI-driven approaches, however, are increasingly being used to streamline sampling activities and cut waste, with some estimates suggesting that no more than 25 percent of fabric is wasted when AI is deployed.
The act of rendering a 2D sketch into the precise technical patterns needed to create a 3D garment has long been a vexing element of the design phase. But AI-powered pattern intelligence platforms can now combine new sketches with a brand’s existing library of patterns and construction techniques on the one hand, and real manufacturing intelligence on the other, turning sketches into product-ready designs at a fraction of conventional times and costs. When used effectively, AI can accelerate pattern-making by up to 70 percent. Next, brands need to make physical samples to test these patterns and their fit and finish. Typically, large quantities of fabric are cut, discarded or altered at this stage of garment testing and pattern modification. AI-driven approaches, however, are increasingly being used to streamline sampling activities and cut waste, with some estimates suggesting that no more than 25 percent of fabric is wasted when AI is deployed.
More broadly, the overproduction of garments is the primary driver of fabric waste. AI’s strength at predictive demand forecasting is tackling this challenge head-on, revolutionising planning, helping brands produce output volumes that near-accurately match demand for a given time of year, and saving millions of tons of fabric that would otherwise have been wasted. Fashion giant Zara, for example, works with an AI partner to analyse more than 3 million social media images every day in order to forecast seasonal or period-specific trends. Drawing on its findings, Zara ensures that 85 percent of its production takes place in-season and that static inventory is avoided.
Production
AI is also shaping production itself. Garment factory floors are witnessing the emergence of a new category of AI, popularly called ‘physical AI’, that interacts with materials and makes real-time adaptations. This is a novel form of automation, where high-tech cameras and sensors feed data to AI systems, allowing the latter to perform a “sense, think, act, learn” feedback loop.
Beyond defect detection, physical AI can also “analyse fabric properties dynamically” and use this knowledge to “optimise cutting patterns in real time”, rather than have factories run inflexibly with preset patterns that result in large volumes of fabric offcuts, or bits and pieces with no practical use.
These systems mark a radical departure from traditional quality control processes that identify defects and recognise an item as waste only after its production is complete, and much material, labour, and energy have already been expended. Physical AI spots defects as they occur, and early diagnosis prevents a cascading waste of resources. Beyond defect detection, physical AI can also “analyse fabric properties dynamically” and use this knowledge to “optimise cutting patterns in real time”, rather than have factories run inflexibly with preset patterns that result in large volumes of fabric offcuts, or bits and pieces with no practical use. Taken together, physical AI and more traditional AI-driven automation systems are having a considerable impact on waste reduction during production, and are likely to be instrumental in reducing the estimated 92 million tons of textile waste produced globally every year.
Post-production
AI deployment during the post-production phase takes a variety of forms. Arguably, the most frequently discussed of these is AI’s role in optimising transport, logistics, and warehousing. This triad of functions is by no means unique to the fashion industry, but finding newer AI-enabled ways to streamline them could go a long way towards reducing the industry’s carbon footprint.
A difficulty that has traditionally plagued online fashion retailers is the inability of customers to try on clothes before buying them, resulting in high rates of product returns and a drag on profitability. In a bid to address this gap, several AI startups have begun to specialise in providing ‘try-on’ technology. The startup Catches, for example, has built a platform that lets users create a realistic digital twin to try on clothes virtually, and the solution is being used by luxury brand Amiri and others. Similarly, Shopify has integrated startup Genlook’s AI try-on app with its e-commerce platform. The cumulative effect of innovations like these does appear to be a rise in conversion rates and a reduction in returns, and a consequent reduction of emissions associated with returns.
AI can identify and distinguish fibres more quickly and accurately than ever before, and this has led directly to better sorting, faster recycling, and lower processing costs.
Finally, at a time when fashion and textile waste has reached crisis proportions—85 percent of all clothes are discarded annually, and only 1 percent of material used to produce clothing is recycled into new clothing—the use of AI for recycling is a game-changer. AI can identify and distinguish fibres more quickly and accurately than ever before, and this has led directly to better sorting, faster recycling, and lower processing costs.
Levi’s’ Labour Lost?
Even as AI makes the fashion industry more sustainable, the human costs of its deployment are troubling. The oft-made observation that AI, unlike earlier forms of automation, will impact blue- and white-collar workers alike, is now becoming distressingly evident across the sector.
In April 2026, videos of workers at an Indian garment factory stitching fabric wearing head-mounted cameras sparked outrage. Speculation was rife that the cameras were being used to collect data to train AI systems that would ultimately replace the workers. In neighbouring Bangladesh, the world’s second-largest garment exporter, around 60 percent of apparel workers or 2.7 million people, could lose their jobs to automation, including AI.
The jobs of knowledge workers across the fashion sector are at risk, too. Projections indicate that AI could eliminate 60–80 percent of fashion design and development jobs by 2028. Not unpredictably, pattern designers, junior and technical designers will be hardest hit, while creative workers in high-end luxury and couture may remain relatively protected.
Speculation was rife that the cameras were being used to collect data to train AI systems that would ultimately replace the workers.
The tradeoff is a tough one. Many of the points at which AI is advancing sustainability and circularity are also functions where it stands to make human interventions redundant. Yet, as companies look to economise and attain greater efficiency, there is little doubt that the march of AI will continue.
New pathways for the restitution and reskilling of factory workers may need to be explored. For instance, a percentage of the efficiency gains from AI-automated lines could be diverted into a national transition fund for displaced workers, and direct financial restitution could be considered for veteran workers for whom it may be too late in their careers to pivot to other roles.
For designers, technical developers, and other creative personnel, upskilling should not involve trying to compete with AI on speed, but to master it as a high-level ‘creative director’. This approach ought to be mainstreamed into fashion courses, design schools, and technical training curricula, as it will come to define the future of fashion.
Anirban Sarma is Director of the Centre for Digital Societies at the Observer Research Foundation
[1] The H&M Group is a global conglomerate of fashion brands and businesses.
[2] Namrata Rana and Utkarsh Majumdar, Balance: Responsible Business for the Digital Age (New Delhi: Westland, 2018)
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! to access our curated content — blogs, longforms and interviews.
