Disconnected data systems and skills shortages are the two main barriers stopping textile manufacturers from cutting waste with artificial intelligence, according to a new industry white paper reported by Ecotextile News from Zurich.

For sourcing and production teams, the finding lands at a practical level. AI tools for waste reduction — whether for fabric utilisation, defect detection or demand-matched production — depend on clean, connected data streams across mill, factory and brand systems. Where that plumbing doesn't exist, the technology simply can't be deployed, however strong the vendor pitch.

What's holding adoption back?

The white paper identifies two structural problems:

• Data fragmentation. Production, quality and inventory data sit in separate systems that don't talk to each other, so AI models can't be trained or run on a complete picture of the operation.
• Skills shortages. Mills and factories lack the in-house expertise to implement and maintain AI tooling, and the sector competes for that talent against industries with deeper tech budgets.

That combination matters because waste reduction is one of the few AI use cases with a clear return for textile suppliers: less offcut, less overproduction, fewer seconds. The barrier isn't the business case. It's the infrastructure and the people to run it.

Who does this concern?

Primarily tier-two and tier-three suppliers — mills, dyehouses and component makers — where digitisation has historically lagged behind cut-and-sew tier-one factories. Brands pushing digitised traceability and material efficiency targets down their supply chains should read the finding as a warning: suppliers can't deliver AI-driven efficiency gains on systems that were never integrated in the first place.

The white paper's framing also carries a compliance angle. Waste and material-efficiency claims increasingly trace back to regulation — the EU's circular economy framework and textile strategy both push producers toward measurable reductions. If the data needed to run AI waste-reduction tools is fragmented, the data needed to evidence compliance improvements is fragmented too.

What should sourcing teams do?

The decision this forces is sequencing. Before any AI procurement conversation, brands and their suppliers need to audit whether production data is standardised and connectable across the relevant tiers. Where it isn't, the near-term investment is integration and training, not models.

The white paper, as reported by Ecotextile News on 16 October, doesn't name specific suppliers or brands — so treat any vendor claim of easy AI-driven waste cuts as unproven until the underlying data architecture is on the table.

The bottleneck isn't the algorithm. It's the plumbing.