The WTO's Digital Trade Rules Are Ready. Are Small Businesses?

The rulebook caught up. The capability did not.

For most of the last decade, the story told about small firms and global commerce has been a story about rules. Remove the tariffs on digital goods, standardize electronic customs, recognize e-signatures across jurisdictions, and a workshop in one country can, in principle, sell to a buyer in another as easily as a multinational can. Much of that rulemaking has now happened. The question this piece examines is whether the policy progress translates into practical access, and the answer, drawn from the WTO’s own material, is that it does not automatically do so.

This is not a criticism of the rulemaking. Tariff-free digital transmission and paperless trade are genuine enablers, and they remove real friction. The point is narrower and more useful: the constraint that keeps a micro or small enterprise from actually trading across borders has moved. It now sits below the policy layer, in the operational capability a firm needs to comply, ship, get paid, and be found. That layer is being addressed far more slowly, and by very different actors, than the treaty text.

What the digital trade rules actually settled

Two instruments carry most of the weight in the WTO framing of small-business digital trade. The first is the moratorium on customs duties on electronic transmissions, which keeps cross-border digital deliveries, software, media, design files, and the like, free of tariffs. The second is the Trade Facilitation Agreement, whose provisions on electronic signatures, paperless trading, and electronic invoicing are explicitly presented as reducing the fixed administrative cost that falls hardest on the smallest exporters.

The logic is sound. Fixed compliance costs are regressive by nature: the same customs paperwork that a large exporter amortizes across millions in shipments can be prohibitive for a firm shipping a few thousand dollars’ worth. Digitizing and standardizing that paperwork lowers the fixed cost, which is precisely the cost structure that theory predicts keeps small firms out of trade. On paper, this is the right lever pulled in the right direction.

What the WTO’s own handbook still admits is missing

The most useful evidence here is not a critic’s. It is the WTO’s own. Its handbook for policymakers on supporting micro, small, and medium-sized enterprises names the binding constraints in plain terms even as it frames the moratorium and trade-facilitation provisions as enablers. Small firms, it observes, shoulder disproportionate burdens from non-digital customs procedures, lack familiarity with the mechanics of e-commerce, and face persistent transport and logistics gaps.

Read carefully, that is an admission that the enabling rules and the disabling frictions coexist. The moratorium removes a tariff a small firm may never have been large enough to pay in volume; the customs-familiarity gap, the payments gap, and the logistics gap remain, and those are the frictions that actually decide whether a small firm can fulfill a cross-border order at all. The policy layer is being fixed faster than the operational-capability layer, and the WTO’s Small Business and Trade materials do not pretend otherwise.

The MSME digital divide predates the AI era

If the capability gap were unique to trade paperwork, it might close as digitization spreads. The wider evidence says it is structural and general. Across OECD economies the divide between small and larger firms is widest precisely for the most sophisticated tools, the ones a modern cross-border operation depends on.

What export data reveals about who actually crosses borders

The clearest evidence that open rules do not equal open access comes from the country with among the most developed digital infrastructure. In the United States, small firms make up 97.2 percent of exporting firms, 270,014 of all identified exporters in 2023, yet account for only 33.0 percent of identified-firm export value, some 588.4 billion dollars. Small firms participate in cross-border trade in large numbers and at a systematic per-firm scale disadvantage.

That pattern is exactly what heterogeneous-firms trade theory predicts. Marc Melitz’s 2003 model shows that exporting carries fixed costs only above-threshold-productivity firms can profitably bear, so the firms that self-select into cross-border trade are systematically different from those that do not. Lowering the fixed cost, which is what the WTO’s trade-facilitation agenda does, should let more firms clear the threshold. But the threshold has never been only tariffs and paperwork. It also includes the capability to be found, trusted, and transacted with by a buyer who is not in your town, and that capability is not distributed by treaty.

The broadband floor beneath the rulebook

It is worth stating the least glamorous constraint plainly, because it sits under every other layer. A firm cannot use paperless customs, digital payments, or any of the machine-legible discovery the modern buyer relies on if it cannot reliably get online. As of the FCC’s 2024 Section 706 report, roughly 24 million Americans, about 7 percent of the population, lacked access to fixed 100/20 Mbps broadband, rising to nearly 28 percent of rural Americans and more than 23 percent of people on Tribal lands. A separate nationally representative survey found about 20 percent of rural small businesses were not using broadband at all.

This is the floor beneath the ceiling. The digital trade rules assume a connected firm. For a meaningful share of the smallest and most rural businesses, that assumption does not hold, which means the enablement is real but conditional, and the condition is unevenly met.

The newest layer opens the door before the leverage exists

The pattern is not a relic of an earlier era; it is repeating in real time. Agentic commerce, where an AI assistant completes a purchase on a buyer’s behalf, is arriving as a new distribution layer before small merchants have meaningful leverage inside it. OpenAI’s Instant Checkout, built on the jointly developed Agentic Commerce Protocol, launched to United States ChatGPT users in February 2026, and a payments provider has separately positioned its own protocol server to onboard, in its stated ambition, tens of millions of small businesses.

The reported reality at launch was narrower. A Forrester analyst noted that only around 30 Shopify merchants were actually live on Instant Checkout in that same month, and OpenAI shifted in March 2026 toward a more merchant-controlled checkout model, an implicit concession that the initial agent-led design under-served merchant control and visibility. This is a live, fast-moving 2026 story and the adoption figures are press-sourced rather than audited, so they should be read as directional. The structural point, however, is the familiar one: the legal and technical door opens first, the practical leverage for the small merchant arrives later, if it arrives.

How to read the gap

The accurate reading resists two easy errors. The first is triumphalism, the claim that open digital trade rules have democratized global commerce for small firms. The evidence does not support it: participation is wide but shallow, the capability gap is structural, and it predates the AI era rather than being caused by it. The second error is fatalism, the claim that small firms are simply locked out. That is equally wrong. The rules are genuine enablers, adoption is rising, and the smallest firms in some categories are closing specific gaps faster than in prior technology cycles.

The defensible conclusion sits between them. Every wave of borderless commerce, globalization, e-commerce, and now agentic commerce, opens the legal and technical door before it opens the operational one, and the capability threshold to walk through re-forms at each new layer. For a small business, the practical implication is not to wait for the rules and not to trust that open rules mean open access. It is to build the specific, unglamorous capability the current layer demands, being connected, being compliant, and being legible to the systems buyers now use to find and compare suppliers, because that capability is what the treaty text has never supplied.