The fashion industry is notorious for its carbon footprint, responsible for around 10 per cent of global emissions according to the United Nations Environment Programme. For years, the fashion industry has made grand promises about sustainability. Brands tout their commitments to net-zero emissions, circular fashion, and ethical supply chains, yet the reality remains unchanged: overproduction is accelerating and emissions are rising. Now artificial intelligence (AI) has entered the conversation, heralded as a game-changer for sustainable fashion. The narrative is seductive: AI can fine-tune production, minimise overstock, and streamline demand. However, this argument ignores a critical issue: AI’s role in fueling consumption. A fundamental economic principle at play here is Jevons’ Paradox, the idea that technological advancements meant to improve efficiency often leads to increased consumption. The fashion industry’s AI adoption is an example of capital-biased technological change, where investment in technology increases productivity but does not necessarily translate into a reduction in resource consumption. Instead, firms reinvest these efficiency gains into scaling production, reinforcing fast fashion’s exploitative business model. The result? Lower marginal costs per unit, higher volumes, and a race to capture market share. Pakistan’s role as a supplier for brands like Zara and H&M means that AI-driven demand could intensify production pressures In economic terms, AI in fashion functions as a positive supply shock: lowering production costs and increasing output. But unlike traditional supply shocks that benefit consumers through lower prices, this one comes at an environmental cost. Take Zara, which has been using AI as a tool to optimise inventory management and prevent overproduction. On paper, it sounds like a breakthrough that will ensure that only what’s needed is produced. But in practice, this efficiency doesn’t slow production or energy use; it accelerates both. Zara’s AI isn’t working to curb fashion’s footprint, it’s working to ensure there are always new styles and trends, and always more reasons for consumers to buy. Shein and Temu are poster children of algorithmic fast fashion as they churn out thousands of new products daily based on real-time consumer demand. During a retail business conference last year in Berlin, Shein claimed that more than 5,000 of their suppliers recently gained access to an AI software platform to analyse customer preferences. The result: more consumption, more landfill-bound fashion. This efficiency in fashion will particularly hit the developing world. Pakistan’s textile sector, which nearly accounts for 60pc of total exports, may soon be at a crossroads because of AI’s increasing role in the fashion supply chain. As global brands push for AI integration to cut costs, Pakistan’s reliance on labour-intensive production could backfire, leading to lower wage growth and rising unemployment in the sector. Pakistan’s role as a major supplier for brands like Zara and H&M means that AI-driven demand could intensify production pressures. This contradicts sustainability goals, as factories are forced to chase tighter deadlines, increasing reliance on unsustainable energy sources like fossil fuel. AI-driven reshoring trends, where brands relocate manufacturing closer to consumer markets, could reduce Pakistan’s competitive edge in textile exports. Countries with advanced automation, like Bangladesh’s smart textile hubs, may outperform Pakistan unless the industry pivots to high-value sustainable textile innovation. Seemingly, AI is not solving fashion’s sustainability crisis, and Pakistan’s textile sector could become trapped in the low-cost, high-volume cycle of AI-optimised fast fashion. At present, Pakistan lacks the regulatory framework needed to address the AI-led transformation in textiles. While the European Union’s Digital Product Passport and New York’s Fashion Act are forcing brands to rethink their supply chains globally, Pakistan’s policymakers have no clear AI or sustainability guidelines for the fashion-textile sector. This leaves the country vulnerable to international brands outsourcing to Pakistan but using AI for sustainability storytelling (greenwashing) while still demanding unsustainable production rates. Without binding environmental laws, local manufacturers will remain under pressure to meet these demands. For AI to genuinely contribute to sustainable fashion, governments must step in with clear but ‘mandatory’ regulations that curb its role in exploitative supply chains. Secondly, Pakistan must invest in high-value textile innovation that values organic and regenerative fibre production to remain competitive in a world where sustainability standards are tightening. Trade policies should be restructured to reward ethical production rather than reinforcing the low-cost- high-volume cycle of AI-optimised fast fashion. Without these measures, AI will not reduce fashion’s environmental impact; it will only accelerate the crisis while deepening economic vulnerabilities for developing nations like Pakistan. The writer is a climate policy analyst and writer and founder of Climate News Pakistan Published in Dawn, The Business and Finance Weekly, March 10th, 2025