AI Product Passports™ — powered by OPEN.MODE

UI mockup of an AI Product Passport by OPEN.MODE. Features a model in a black gown, 94% verification confidence, and 100% Viscose material info. Sleek floating cards and chrome spheres illustrate digital fashion authentication and transparency.

Making Products Machine-Readable for Agentic Commerce

White Paper by OPEN.MODE Innovations Ltd, London, United Kingdom Published: January 13, 2026.

Executive Summary

AI agents are becoming the new shoppers. From personal styling assistants to automated procurement systems, software — not humans — will increasingly decide what gets bought and from whom. For two decades, brands optimised for search engines: keywords, backlinks, schema markup. If you weren't on page one of Google, you didn't exist. That era is ending.

In agentic commerce, the consumer of information is no longer a search engine — it's an AI agent. And AI agents don't crawl web pages. They query structured data, verify credentials, and compare claims programmatically. The optimisation target is no longer your website — it's your Digital Product Passport. The practice is no longer SEO — it's GEO: Generative Engine Optimisation. And the consequence of non-optimisation is the same as it was in 1998, just updated: if your DPP isn't agent-readable, you're invisible to agentic commerce.

Infographic comparing Era 1 Web Search 1998 to 2020 and Era 2 Agentic Commerce 2025 plus. It shows the shift from SEO targeting Google to GEO targeting AI Agents. Non-optimised Digital Product Passports make brands invisible to agentic commerce.

The shift from SEO to GEO — how optimisation targets, practices, and consequences are changing as AI agents replace search engines as the primary consumer of product information.

This paper proposes AI Product Passports — a machine-readable infrastructure layer that makes product claims structured, verified, and queryable by AI agents. We introduce a four-level Assurance Stack (A0–A3) that gives AI agents calibrated confidence in every claim, from self-declared data to independently audited evidence. For fashion brands, the strategic window is now: EU regulation (ESPR) demands compliance by 2027, and the brands that build machine-readable product identities first will be the ones AI agents recommend.

Table of Contents

I. The Shift: From SEO to AI-Led Commerce

For twenty years, brands optimised for humans. They built websites, funnels, SEO strategies, and paid acquisition engines.

Now the buyer is changing.

AI systems like ChatGPT, Gemini, and emerging AI shopping agents are increasingly mediating discovery, evaluation, and purchasing decisions. Instead of browsing, customers will say: "Buy me a long black evening dress under £500, based on my taste you already know, I have a gallery opening next week in New York."

Smartphone interface showing an AI "Shopping Agent" assisting with a black evening dress search. It displays a "Top pick" image of a model in a sculptural gown with design options like "Minimalist neck" selected to refine the curated results.

AI agents will shortlist products before humans ever see them. This creates a structural shift: if AI cannot understand your product, it cannot recommend it.

The Challenge Today

Most product data today is written for marketing, not machines. It lives in PDFs, ERP systems, and product pages — fragmented, unstructured, and impossible for AI to reason over.

At the same time, regulation (ESPR - Ecodesign for Sustainable Products Regulation, Digital Product Passport frameworks) is pushing brands toward traceability and structured reporting. But compliance alone does not equal discoverability. And discoverability alone does not equal trust.

Brands now face a triple challenge:

Timeline of Regulatory, Market, and Strategic tracks from 2024 to 2030. Diamond Collision Points highlight strategic decisions for Battery DPP and Agentic Commerce in 2027, marking a 2025-2027 window to position, invest, and capture market.

Three colliding timelines creating friction in fashion supply chains: EU regulatory mandates (ESPR) assuming machinereadable data that does not exist; market adoption of AI agents demanding verification infrastructure that lags behind; and strategic investment treating DPP as compliance cost rather than competitive capability. The collision—not convergence—creates urgency for organisations to architect DPP systems with agent accessibility as a design principle.

The triple challenge

Regulatory readiness

Can you report what regulators require?

AI readiness

Can machines parse your product data?

Verification readiness

Can your claims withstand algorithmic scrutiny?

Very few brands are prepared for any of these. Almost none are prepared for all three.

II. What Is an AI Product Passport™?

An AI Product Passport™ is a machine-readable product profile that helps AI understand, compare, and recommend your products accurately.

What's the Difference Between a Digital Product Passport and an AI Product Passport™?

A Digital Product Passport (DPP) meets regulatory requirements. An AI Product Passport™ goes further — it makes your product discoverable, comparable, and recommendable by AI shopping agents. Compliance is included.

To understand why that distinction matters, consider something you already trust every day.

The Nutrition Label Analogy

When you buy a granola bar, you look at the nutrition label. You see "10g of protein" and you believe it. You don't call the factory. You don't test the bar yourself. You trust the number because a regulator certified the process behind it.

This works because you're human. You make one purchase at a time, and "good enough" trust is good enough.

Now imagine an AI shopping agent making 10,000 decisions per second — comparing materials, verifying certifications, cross-referencing supplier claims across hundreds of brands simultaneously. It can't rely on "good enough." It needs data it can parse, compare, and verify programmatically.

A Digital Product Passport is the nutrition label. It lists what's required. An AI Product Passport is what happens when the reader is no longer human.

The Oracle Problem

But the analogy has a limit — and this limit is the core problem our research addresses.

A nutrition label assumes the numbers on it are true. In global supply chains, that assumption breaks down. A company can claim their factory is carbon-neutral. A blockchain can prove that claim was recorded. Neither can prove that the factory actually turned off the machines.

This is what our research calls the Oracle Problem: digital assertions about physical products cannot be fully verified through digital means alone. A blockchain can prove a claim was recorded. It cannot prove what happened in a field 8,000 kilometres away.

An AI Product Passport is built around this reality. Rather than pretending all claims are equal, it makes verification depth explicit — every claim carries a confidence level that tells the AI agent exactly how much to trust it, and why.

"A nutrition label tells you what's in the box. An AI Product Passport proves it, shows you the receipts, and tells you exactly how much you should trust the person who wrote the receipt."

This is what the Assurance Stack solves — a four-level framework that gives AI agents calibrated confidence in every claim, from self-declared marketing statements to independently verified physical evidence.

AI Product Passports turn product traceability into commercial infrastructure — without overstating what that infrastructure can guarantee.

What an AI Product Passport™ Does

  • Structures product attributes (materials, origin, certifications) for machine readability

  • Encodes functional and contextual metadata

  • Associates every claim with its evidence basis

  • Applies confidence scoring calibrated to verification depth

  • Enables query-based discovery by AI agents

Three Forces are Converging

Compliance

Digital Product Passport mandates are arriving (EU ESPR) from 2027

Agentic Commerce

AI shopping agents are moving from prototype to deployment

Verification demand

Consumers and regulators reject unsubstantiated claims

III. Commercial Impact and Strategic Opportunity

What This Means in AI-Mediated Commerce

As AI agents increasingly mediate purchasing, products will be:

  • Filtered before human visibility

  • Ranked based on claim clarity and verification depth

  • Deprioritised if claims are ambiguous, unstructured, or unverifiable

Brands That Adopt AI Product Passports™ Can:

  • Increase discoverability in AI shopping interfaces

  • Improve conversion through transparent, graduated trust signals

  • Reduce regulatory risk (ESPR compliance as a byproduct, not an add-on)

  • Turn compliance investment into commercial infrastructure

This is not a marketing layer. It is a structural layer.

A live AI Product Passport: materials, origin, confidence scoring, and designer intent — structured for both human browsing and AI agent queries.

The Specification Cost Advantage

Our research identifies a new cost category emerging in AI-led commerce: specification costs — the investment required to articulate product attributes in machine-interpretable form.

Today, these costs are high. Structuring product data for AI readability requires effort across teams, systems, and suppliers. But specification costs are front-loaded: brands that invest now face declining marginal costs as AI commerce scales. Those that wait accumulate specification debt — an increasingly expensive gap between their product data and what AI systems require.

Early movers build infrastructure. Late movers build workarounds.

Value Across the Supply Chain

AI Product Passports™ do not only benefit brands. Our analysis suggests that mid-tier suppliers (Tier 1–2) may capture disproportionate value by investing in structured, higher-assurance product intelligence. Suppliers who can demonstrate A1/A2 verification gain algorithmic visibility that generic compliance cannot match.

This creates a differentiation window — one that closes as standards mature and verification becomes commoditised.

What We Are Not Claiming

Intellectual honesty matters. We recognise that:

  • The Oracle Problem is structural. Full physical-digital verification remains economically prohibitive for most product categories

  • Most sustainability claims in fashion today sit at A0–A1. That will not change overnight

  • Algorithmic homogenisation is a real risk — when agents rely on similar models, market diversity can narrow

  • AI Product Passports™ are infrastructure, not a silver bullet

We believe the framework that makes its limitations explicit is more trustworthy — and ultimately more commercially durable — than one that hides them.

The Oracle Problem in fashion supply chains: from impossibility to governance. Panel A illustrates the structural gap between digital assertions and physical reality: raw materials (T4; supply chain tiers range from T1: assembly to T4: raw materials) from multiple sources converge at aggregation nodes where exact provenance is converted to probabilistic mass-balance attribution— a transformation no digital system can reverse. Panel B presents the governance response: deterrence (balance scale) makes dishonesty irrational even with low audit probability; belief updating (evidence funnel) treats physical truth as a latent variable refined by sparse verification; and the Assurance Stack (A0–A3) maps to graduated audit probabilities (icon arrays). Critically, the framework does not solve the Oracle Problem but governs decisions under it by encoding evidence provenance and calibrating decision error rates on observable evidence streams rather than asserting verified truth.

Conclusion: The Window Is Now

The brands that win in agentic commerce won't be the ones with the best marketing copy. They'll be the ones whose products AI agents can actually find, read, and trust.

EU regulation (ESPR) sets the compliance deadline at 2027. AI shopping agents are moving from prototype to deployment today. The gap between these two timelines — right now, 2025 to 2027 — is the strategic window. Brands that build machine-readable product identities during this window become the default recommendations. Brands that wait become invisible.

This isn't a technology problem. The infrastructure exists. This is a timing problem — and the clock is already running.

Three ways to move:

  1. Read the full research — the complete framework, all figures, and the technical appendix please contact us below. We will share upon a NDA.

  2. Talk to us — if you're a brand, retailer, or platform exploring AI-readiness for your product data, we'd like to hear what you're building toward. Please fill out this from if you are interested in a pilot or contact us below.

  3. Follow the work — we publish ongoing research, implementation patterns, and industry analysis. Connect with us on LinkedIn to stay in the loop.

Get Involved

We are currently implementing AI Product Passports™ with selected fashion and lifestyle brands.

Brands

Assess your product data's AI-readiness

E-Commerce

Pilot an AI Product Passport™ across your catalogue

Industry Experts

Explore how your supply chain performs under algorithmic scrutiny

FAQs

What is an AI Product Passport™?

1

A machine-readable product profile that helps AI understand, compare, and recommend your products accurately.


How do AI Product Passports increase sales?

2

When AI understands your products, it recommends them to the right customers — increasing visibility, trust, and conversion.


Why does AI matter for e-commerce now?

3

AI is replacing search and ads as the main way people discover products. If AI can’t read your data, it won’t surface your products.


Do I need to rebuild my tech stack?

4

No. OPEN.MODE is designed to work alongside your existing systems — no re-platforming required.


What's the difference between a DPP and an AI Product Passport?

5

A Digital Product Passport meets regulatory requirements. An AI Product Passport goes further — it makes your product discoverable, comparable, and recommendable by AI shopping agents. Compliance is included.