Reinforce, Don’t Reopen: Why Digital Trade Matters in the 2026 USMCA Review

Introduction

When the U.S.–Mexico–Canada Agreement (USMCA) entered into force in 2020, its digital trade chapter was among the most advanced in any trade agreement. Chapter 19 established rules for how information moves across borders, prevented governments from forcing companies to store data locally, and barred them from demanding proprietary software code as a condition for doing business in the region. For the first time, the rules underpinning North America’s digital economy were written down and enforceable. Nearly six years later, the world those rules were designed for has changed.

Cross-border digital trade in North America now exceeds $250 billion annually. Data flows underpin manufacturing, financial services, logistics, and critical infrastructure across the three USMCA economies. While other trade agreements include digital provisions, none operate within a production ecosystem as deeply integrated as North America’s, where goods, services, and data cross borders multiple times before reaching a final consumer. These rules get tested daily through supply chains, transactions, and investment decisions. This is North America’s digital advantage.

That advantage, however, is under strain. Artificial intelligence (AI) has rapidly reshaped the economics of data almost overnight, raising questions about governance, competition, and labor displacement that no trade agreement can currently answer. Mexico is expanding state-led digital infrastructure. Canada is reassessing its reliance on U.S. technology in the wake of trade tensions. And the United States is using export controls and diplomatic leverage to consolidate its position at the center of the global digital ecosystem. While these approaches differ, they reflect a shared trend: greater state control over digital infrastructure and data.

The central question is whether USMCA’s digital chapter can absorb these pressures—technological, geopolitical, and strategic—without losing coherence. North America already possesses the rules, integration, and infrastructure to lead the global digital economy. If the three governments treat the digital chapter as a secondary priority during the 2026 review, they will have let the moment pass to lock in that advantage.

What Chapter 19 Got Right

The USMCA’s digital chapter has delivered genuine economic value. Its core provisions remain sound, and preserving them should be the starting point for the 2026 review. For Mexico and Canada, the prohibition on forced data localization was transformative, allowing firms to plug into regional cloud infrastructure without building costly local data centers. For U.S. firms, the chapter gave them a single legal framework to serve customers across all three markets. As investment in AI infrastructure, data centers, and cloud capacity accelerates across North America, that framework becomes increasingly valuable.

The chapter’s core commitments have aged well. The pressures it now faces reflect not failures of USMCA but challenges that arrived faster than any trade agreement could have anticipated. The priority for the review should be straightforward: Preserve the core and build around it.

The Sovereignty Trap

Governments worldwide are debating “digital sovereignty.” The concern is straightforward: In a world where cloud infrastructure, operating systems, and AI models are controlled by a handful of U.S. and Chinese firms, dependence on either can be exploited. If political relations deteriorate, a government could find its critical digital systems subject to sanctions, service interruptions, or coercion—what experts now call the “kill switch.” China has repeatedly used economic pressure against countries that cross its political interests, from Australia to Lithuania to Canada. And the tariff war between the United States and its neighbors has shown that even its closest allies are not immune. Canadian provinces have threatened to cut U.S. technology firms from public procurement.

The responses have been swift. Mexico’s Digital Transformation Agency is constructing a national digital stack. Canada’s procurement strategies increasingly favor domestic providers. And at the G20 and the Organization for Economic Co-operation and Development (OECD), the conversation has shifted from openness to national control over digital infrastructure. Given what Canada and Mexico have experienced over the past year, these responses are understandable. But as an industrial strategy, they run into a wall.

No country, not even the United States, can produce a fully self-contained national digital stack and remain competitive. The technology supply chain is inherently dispersed: Chips designed in the United States are manufactured in Taiwan with Dutch and German machinery, then assembled with South Korean and Japanese components. No single country controls the full chain or can replicate it affordably. At the scale required to run a modern economy, there are only two builders of digital infrastructure capable of operating globally: the United States and China. If pursued unilaterally, national stack ambitions will raise costs, weaken security, and fragment the region from within.

The global digital economy is bifurcating into two ecosystems. One is commercially developed, interoperable, and operates across borders, anchored by U.S. platforms but built on globally dispersed hardware and talent. The other is China’s: closed by design, governed by state control, and steadily expanding its reach through infrastructure investments and financing across the developing world. For countries currently outside China’s system, the choice is not between sovereignty and openness. It is between participating in the interoperable ecosystem, with all its imperfections and asymmetries, or watching fragmentation push governments toward the only alternative that operates at scale. But that alternative comes with conditions: state surveillance built into the architecture, political leverage embedded in the financing, and dependency on a system designed to serve Beijing’s strategic interests, not its users’.

North America already has something better. The United States has the platforms, hyperscale cloud, and AI models. Mexico has a growing nearshoring-linked IT workforce and expanding digital and tech exports. Canada has world-class AI research and energy generation capacity. Together, within an integrated production ecosystem where goods, services, and data already cross borders continuously, they form a combination no single country can replicate, and no other trade bloc has locked in through enforceable rules. Every dollar spent building a redundant national system is a dollar not invested in the regional infrastructure that already works. The sovereignty impulse is understandable; the math behind it is not.

The United States has the leverage to set the terms of North America’s digital future. But leverage alone will not hold the bloc together. It needs to offer concrete commitments: guaranteed continuity of digital services so the kill switch concern has an answer, reciprocal access to procurement markets, aligned regulatory frameworks, and enforceable guardrails against the discretionary use of technology access as a tool of coercion. The strongest digital bloc in the world is not built on dominance. It is built on the confidence that the rules apply to everyone.

Mexico’s Regulatory Overhaul and the Credibility Gap

Over the past two years, Mexico has overhauled the institutional architecture that underpins its digital economy. The independent telecommunications regulator, the Instituto Federal de Telecomunicaciones, was dissolved and replaced by a new body housed inside the Digital Transformation Agency, which operates at the ministry level under the executive branch. Other autonomous bodies responsible for competition and data privacy were similarly absorbed into the executive. Mexico has put the referee in a home team jersey and asked investors, foreign or domestic, to trust the calls.

The problem is that the USMCA rulebook was written for independent referees. Article 18.17 requires each country’s telecommunications regulator to be structurally separate from any supplier of public telecommunications services and to hold no financial or management stake in one. A footnote specific to Mexico goes further, requiring that the regulator be autonomous from the executive branch and independent in its decisions. Whether Mexico’s new institutional architecture can meet those standards remains untested. Whether the new telecoms regulator can exercise genuine independence within a hierarchy that reports to political leadership, particularly on questions involving state-owned competitors, will be an early and visible test.

The competition landscape compounds the concern. Mexico’s revised competition law now includes built-in exceptions for state-owned enterprises. This is directly relevant to CFE Telecomunicaciones e Internet para Todos (CFE TEIT), a state telecom operator that benefits from subsidized spectrum and public financing and now competes directly with private operators in retail services. If the new competition authority cannot apply the same rules to state-owned and private firms, the level playing field that the USMCA’s telecommunications and competition chapters require will exist on paper but not in practice.

Article 30B of the Fiscal Code deserves particular attention. It requires companies to provide permanent, real-time access to platform data, including income and user information, for fiscal purposes. The provision enters into force on April 1, 2026, just three months before the USMCA review deadline, and while framed as a tax compliance tool, it raises concerns that extend well beyond revenue collection.

The law states that access will be limited to fiscal purposes, but it establishes no published proportionate safeguards, no independent oversight mechanism, and no clear boundaries on secondary use. No formal localization mandate is required when regulation produces the same outcome by other means. For foreign firms, permanent government access to operational data, without clear limits or judicial oversight, fundamentally alters the risk calculation for operating in Mexico. Whether Article 30B is compatible with Chapter 19’s cross-border data flow commitments is an open legal question, and one that the review will almost certainly surface.

Mexico is not the only country where digital trade commitments face domestic pressure. Canada’s Online Streaming Act and its short-lived digital services tax (DST) raised their own questions about compatibility with USMCA provisions before the DST was withdrawn under U.S. pressure. But Mexico’s changes are structural. When the institutions responsible for enforcement lack independence, when competition rules exempt the state’s own players, and when new data access obligations arrive without safeguards, the cumulative effect is not a single violation but a steady accumulation of risk that makes investment decisions in Mexico harder to justify. That is Mexico’s credibility gap. It weakens not only investor confidence but Mexico’s own negotiating position: Every unresolved concern over regulatory independence or data access becomes leverage for Washington to extract concessions elsewhere in the review. And no amount of progress on autos, steel, or other headline items will make it go away.

The Elephant in the Room: Data, Trade Balances, and the First Data State

The economist Dan Ciuriak, formerly deputy chief economist at Global Affairs Canada, has offered a useful thought experiment for understanding a dimension of the USMCA that none of the three governments have seriously reckoned with: the value of data in North American trade.

Every time a consumer in Mexico or Canada uses a U.S. search engine, social media platform, or cloud service, they generate data that feeds algorithms, trains AI models, and drives advertising revenue. That value accrues overwhelmingly to U.S. firms. But in trade statistics, these exchanges register as zero: free services delivered in exchange for data that has no recorded price.

Ciuriak estimates that if the value of data were reflected in trade flows, it would add $800–$900 billion to annual U.S. exports. Even if the precise figure is debatable, the direction is not. McKinsey estimates that cross-border data flows contribute $2.8 trillion to global GDP, and the OECD warns that restricting them could cut global output by up to 5 percent. The United States captures a disproportionate share: It hosts nearly half the world’s data centers, and the revenue generated from cross-border data accrues overwhelmingly to U.S. firms. By any reasonable measure, the United States is the world’s first “Data State.”

The implications cut across North America. For Canada, which appears to run a goods trade surplus with the United States, accounting for data flows would likely flip the balance to a substantial deficit, undermining the narrative that Canada benefits at U.S. expense. For Mexico, the same logic applies: The value that U.S. platforms capture from Mexican consumers and firms through data is enormous and entirely unrecorded.

None of this means Washington will revise how it calculates trade deficits. Given the administration’s preference for narrowly defined goods deficits to justify its tariff posture, that is not realistic. But the insight matters strategically. Mexico and Canada enter the review under pressure to make concessions, in part because the deficit numbers frame them as the primary beneficiaries of the current arrangement. The data tell a different story. The digital framework that the three countries built together already delivers its largest returns to the United States.

The strongest case for protecting Chapter 19 is not that Mexico and Canada need it. It is that the United States benefits from it more than any official metric reflects. Free data flows across North America, with value flowing to U.S. platforms. Weakening Chapter 19 could cost Washington more than its partners.

Recommendations: Reinforce, Don’t Reopen

The digital chapter does not need to be rewritten. It needs to be reinforced with mechanisms that address the new landscape without creating new uncertainty. Three concrete steps would accomplish this.

First, take reopening Chapter 19 off the table. The commitments on cross-border data flows, data localization, and source code protection remain the legal backbone of North America’s digital economy. Reopening them would generate exactly the kind of uncertainty the three countries need to prevent.

Second, treat digital trade as an early deliverable, not a leftover. If autos, steel, and energy dominate the review, as they almost certainly will, the digital chapter risks being deferred indefinitely. That would be a mistake. Digital trade is fundamental to the economic future of all three countries. An early agreement would signal that the review is producing results, even as harder negotiations continue.

Third, craft an AI side letter. The three governments should agree on a protocol, attached to the USMCA but outside Chapter 19, that establishes shared principles for AI governance and a framework for regulatory interoperability. Concretely, this could include common AI risk definitions, mutual recognition of safety standards, coordinated approaches to compute infrastructure and data center investment, and a standing trilateral working group to keep the framework current as the technology evolves. The institutional momentum already exists: in March 2026, the Office of the United States Trade Representative and Mexico’s Secretaría de Economía launched bilateral technical discussions in advance of the USMCA Joint Review. An AI side letter would build on that channel and extend it trilaterally, giving all three governments a concrete, forward-looking deliverable to show that North America can still build together, not just fight over what it already has.

Conclusion

Digital trade is no longer one chapter among many. It is the operating layer through which North American integration functions, supply chains communicate, and data flows are sustained. Yet it remains the dimension of the agreement whose value is hardest to see in official statistics and easiest to take for granted at the negotiating table.

North America’s digital advantage is real, but it is not self-sustaining. If the three countries treat the 2026 review as an opportunity to reinforce that foundation, the digital chapter could become one of the agreement’s most durable assets and an early signal of forward progress toward extension. If they neglect it, the costs will compound quietly across every sector that depends on data, connectivity, and trust. And what North America chose not to protect, its rivals will replicate, scale, and use to pull the rest of the world into their orbit.

Diego Marroquín Bitar is a fellow with the Americas Program at the Center for Strategic and International Studies in Washington, D.C.