The single biggest brake on artificial intelligence adoption across the textile and apparel supply chain isn't the technology — it's fragmented, inconsistent data, according to a report from Ecotextile News.

That's a sourcing-desk problem before it's an IT problem. AI tools for everything from fibre traceability to factory-level environmental reporting only work if the data feeding them arrives in consistent formats from consistent points in the chain. In most textile supply chains, it doesn't.

What does 'data fragmentation' actually mean here?

It means the same information — a fabric composition, a mill's water usage, a batch number — gets captured differently at each tier (each step of the chain: fibre, spinning, weaving or knitting, dyeing and finishing, cut-make-trim). One supplier logs it in a spreadsheet. Another uses an ERP system. A third doesn't log it at all.

The result is that brands trying to deploy AI-driven analytics can't build a clean dataset across their supplier base. The model is only as good as the messiest tier-3 (sub-supplier) feed in the chain.

Why this lands hardest on compliance teams

For practitioners, the timing matters. AI has been pitched as the answer to mounting reporting burden:

• EU textile strategy disclosure requirements
• Digital product passport preparation
• Scope 3 emissions accounting
• Due diligence obligations under supply chain laws

All of these depend on primary data from mills and finishing houses. If that data arrives fragmented, AI can't fix it — it just automates the gaps. The Ecotextile News finding suggests the industry's constraint sits upstream of the algorithm: at the point of capture.

What changes for suppliers?

Suppliers should expect the next wave of buyer audits and onboarding questionnaires to probe data practices, not just labour and chemical compliance. A mill that can hand over machine-readable production and environmental data in a standard format becomes measurably easier to work with — and harder to drop — than one that answers in PDFs.

Brands, for their part, face a sequencing decision. Investing in AI platforms before investing in data standardisation across the supplier base means paying twice: once for the tool, again for the cleanup.

The decision this forces

The report's implication is blunt. Before any AI procurement, sourcing and sustainability teams need to agree on what data tiers one through three must capture, in what format, and who verifies it. That's a supplier-development exercise, and it's slower than a software rollout.

The story here isn't that AI has stalled. It's that it's waiting — on the unglamorous work of standardising data at the mill gate.