AI-Driven Design Systems Challenge Ownership of India’s Textile Heritage

Bitupan Deka, an assistant professor of computer technology at Shillong’s MIT University, encountered a troubling trend while shopping for a traditional “gamusa” scarf in Meghalaya, India. He observed that shops repeatedly sold scarves featuring only a limited set of patterns. Upon inquiry, he learned that weavers in Sualkuchi, Assam, were paid a one-time fee of approximately 2,000 rupees (about $21) per design, without royalties. Manufacturers deliberately limited the number of designs to reduce costs.

Motivated by this discovery, Deka developed an AI-based system capable of generating an infinite variety of new textile designs. He clarified that his invention is not intended to replace traditional designers but to serve as a tool enabling them to create modern motifs more easily and offer consumers greater variety. His university’s research and development department is currently applying for a patent on the system, which Deka plans to commercialize as a design service for the textile industry.

India’s textile traditions are deeply rooted in local culture, with patterns and motifs embodying the history, mythology, and environment of their regions. Weavers inherit these designs through generations, using locally sourced colors and handcrafted techniques such as carved clay and wooden frames. Their work reflects a cultural memory and artisanal hesitation that a computer cannot replicate.

However, AI systems like Deka’s learn by analyzing images and generating countless variations, effectively reproducing traditional motifs without the cultural context. A comparable initiative, the Rekh Sanganeri Dataset, focuses on hand-block printed Sanganeri textiles from Rajasthan. It contains high-resolution images annotated to capture imperfections like ink bleed and misalignment, which characterize handmade work. Its creator, Reddy Sai Nivas C, emphasizes that the dataset is a resource for research and art, not a substitute for artisans.

This juxtaposition highlights a paradox: digital tools designed to preserve crafts simultaneously facilitate their replication without involving the original artisans.

Dinesh Chhipa, a member of the Chhipa artisan and merchant caste renowned for dyeing and block printing, and a Bagru designer from Rajasthan, acknowledges the limitations of traditional craft production. He notes that even with substantial financial incentives, mass production of handcrafted pieces is unfeasible. Consequently, he does not fault the use of digital technology and AI to meet demand gaps. However, he insists that the names and identities of regional crafts such as Bagru and Jaipuri should not be appropriated, as these designs originate from specific cultural contexts and community needs.

India has been actively digitizing its textile heritage through initiatives like the Botanical Survey of India’s historical collections and the Bharatiya Vastra Shilpa Kosh, a national portal developed by the National Institute of Fashion Technology and the Centre for Development of Advanced Computing (C-DAC). While such documentation is vital for preserving fragile traditions, it also enables AI systems to access and reproduce these crafts at scale.

The evolving intersection of AI technology and traditional textile crafts in India raises complex questions about intellectual property rights, cultural preservation, and fair economic practices. As AI-generated designs become commercially viable, stakeholders must navigate the challenges of protecting artisanal heritage while embracing innovation.