Artificial intelligence fails to reach MSMEs in apparel industry

Amid China’s push for digital transformation in traditional industries, a new study reveals that micro, small, and medium-sized enterprises (MSMEs) in the apparel manufacturing sector are struggling to integrate artificial intelligence (AI) into their operations. While government policy promotes intelligent manufacturing as a cornerstone of economic modernization, practical barriers are slowing progress for the 92% of firms that make up the backbone of China’s apparel industry.

Published in Sustainability, the study titled “Artificial-Intelligence-Enabled Innovation Ecosystems: A Novel Triple-Layer Framework for Micro, Small, and Medium-Sized Enterprises in the Chinese Apparel-Manufacturing Industry” presents findings from 20 semi-structured interviews with industry managers, academic researchers, and association representatives. The research identifies the required AI capabilities, the barriers to adoption, and proposes a novel three-layer innovation ecosystem to overcome systemic challenges.

What AI capabilities do apparel MSMEs require most?

The study identifies two primary AI capabilities as essential for apparel MSMEs: adaptive production and augmented human–AI collaboration. Adaptive production includes flexible order scheduling, precision forecasting, and efficient quality control - core functions that AI can dramatically improve through automation and predictive analytics. These features are vital for MSMEs that typically operate in low-margin, high-customization environments.

Augmented human–AI collaboration emphasizes symbiotic workflows between skilled labor and AI tools. In many processes, such as pattern cutting, material selection, and production adjustment, human decision-making remains irreplaceable. AI technologies, while capable of enhancing efficiency, do not eliminate the need for human oversight in complex or creative tasks.

Big data analytics also emerged as a critical enabler. By analyzing consumer preferences, supplier performance, and production data, AI helps MSMEs adapt more quickly to shifting market demands. However, these tools require technical infrastructure and skilled personnel, resources that are often in short supply at the MSME level.

What systemic barriers are hindering AI adoption?

Despite the apparent benefits, the study outlines three major categories of barriers hindering AI adoption in China’s apparel MSMEs: industry-level, university-level, and government-level.

At the industry level, firms face high costs, low returns on investment, and complex production demands that limit automation. Many MSMEs lack awareness of AI’s practical applications and are hesitant to invest without clear, immediate benefits. Workforce readiness is another concern; the aging labor force and low technical proficiency create additional resistance to digital upgrades.

University-level obstacles center around talent shortages and misalignment between academic training and industry needs. Most apparel programs focus on design, neglecting the technical aspects of AI-driven production. As a result, graduates are often unprepared to work with smart manufacturing tools, widening the talent gap in AI implementation.

Government-related challenges include uneven policy support, a focus on large enterprises, and weak institutional coordination. While national strategies support AI adoption, regional execution is inconsistent. Many MSMEs, especially those outside major industrial zones like the Yangtze and Pearl River Deltas, are excluded from subsidy programs or lack access to industry associations with the necessary technical expertise.

How can AI ecosystems be built for MSMEs?

To address these issues, the study proposes a Triple-Layer AI-Enabled Innovation Ecosystem Framework, which connects firm-level capabilities, structural barriers, and ecosystem-wide collaboration mechanisms. The model identifies three layers:

• Layer 1: Core AI capabilities like adaptive production and human–AI collaboration
• Layer 2: Barriers stemming from industry, university, and government constraints
• Layer 3: Collaborative mechanisms enabling ecosystem-wide support

Thirteen propositions are outlined, offering actionable strategies to build this ecosystem. For governments, establishing regional innovation hubs and offering MSME-specific funding are essential steps. Public-private partnerships can foster cross-sector collaboration, ensuring that AI initiatives reach beyond large firms.

Universities must reform their curricula to integrate interdisciplinary AI training and develop innovation centers focused on smart apparel manufacturing. Hands-on training programs and collaborative research initiatives are also necessary to close the talent gap.

For industry stakeholders, the report calls for standardized data-sharing protocols and mentorship programs led by AI front-runners. This would ensure knowledge diffusion across the sector, preventing further marginalization of resource-poor MSMEs.