On July 1, 2026, the first joint review of the Canada-United States-Mexico Agreement (CUSMA) opened with the United States declining to commit to a 16-year extension, tipping the pact into annual reviews. For most Canadians, the headlines belong to autos, steel and dairy. But a quieter development deserves attention: digital trade and financial services — two domains the agreement treats as separate — are beginning to merge, and stablecoins are where they meet.
Consider a transaction. A Toronto company licenses software from a Texas vendor and pays in US-dollar stablecoins issued by a Canadian entity under Canada’s Stablecoin Act. The software crossing into Canada is digital trade under CUSMA Chapter 19. The payment crossing the other way is money transmission in the form of a data transfer — a textbook financial service under Chapter 17 — yet, in both directions, only data moved. The software was data; the payment was data. The separation between them exists nowhere but in law and in the institutions whose authority depends on maintaining it. CUSMA assumes that this separation can be read from the item transacted. Increasingly, it cannot.
The agreement assumes digital commerce and finance occupy separate lanes. Chapter 19 protects cross-border data flows, bans data localization and shields source code. Chapter 17 governs financial services and preserves, through the prudential exception in article 17.11, and gives regulators the freedom to act in defence of financial stability. As a result, the drafters built two walls to keep the transacting entity and the product being sold apart. The first wall asks who is transacting: Chapter 19’s protections run to a covered person (a category expressly excluding anyone covered by the financial services chapter), so a bank stays in Chapter 17 and loses Chapter 19’s data-flow protections. Call this the entity wall. The second wall asks what is being sold: Chapter 19’s digital product definition excludes “a digitized representation of a financial instrument, including money,” so a payment instrument cannot claim the same treatment as software. Call this the product wall. Both walls made sense in 2018, when financial functions were performed by financial institutions and money was plainly a different thing from a data transfer. Now, however, they are better understood as surveyor’s marks, showing where the boundary was believed to run.
Turning Ambiguity into Advantage
Stablecoins walk through both walls, and this is no accident. Much of their commercial appeal lies there: monetary function without the institutional form and territorial borders to which regulation attaches. The entity wall assumes financial functions are performed by financial institutions. However, the firms issuing payment stablecoins and operating wallets are not limited to traditional financial institutions and could include technology companies outside Chapter 17’s definitions, so they carry a financial function into Chapter 19 with its protections intact.
The product wall fails twice over. It assumes, first, that money and data are separable. But a stablecoin payment is a transfer of value in the form of a transfer of data; the message is the asset, the payment and the settlement. Second, it assumes a settled taxonomy of financial things, while US law has concluded otherwise. The Guiding and Establishing National Innovation for US Stablecoins Act (GENIUS Act) provides that a payment stablecoin from a permitted issuer is neither a security nor a commodity, and the 2022 amendments to the Uniform Commercial Code place stablecoins outside money altogether, treating them as controllable electronic records — a category created because the existing ones did not fit.
Carving money out of digital products removes the token from Chapter 19’s customs and non-discrimination provisions while leaving the rails largely intact. What is left is an instrument functionally both a financial service and an act of digital trade, yet disciplined as neither. In other words, the drafters fixed the property line by reference to the river, and the river moved.
Neither Canada nor the United States has resolved this at home, and the United States has been the quickest to turn the ambiguity into advantage. The GENIUS Act does more than license issuers. Reserves must sit in cash and short-term treasuries, so every coin issued generates demand for US government debt. Issuers must be able to freeze, seize or burn tokens on lawful order, so American enforcement reach travels with the instrument, and sanctions obligations apply US policy at the level of the instrument itself. With more than US$270 billion in circulation and volumes rivalling the major card networks, US stablecoins in a Canadian resident’s wallet are not simply a payment instrument in a foreign currency. Rather, they are a means through which American monetary, fiscal and enforcement policy operate inside Canada, without any Canadian regulator having agreed to this arrangement, and without any treaty being broken.
Canada responded to this situation in March 2026. The Stablecoin Act in Bill C-15 places fiat-backed payment stablecoins under Bank of Canada supervision, with reserve, redemption and governance requirements expected in force by 2027. Washington’s and Ottawa’s strategies face in opposite directions, with Washington regulating the stablecoin as a product to scale globally, while Ottawa regulates it as a payment instrument inside Canada’s payments perimeter.
What Canada Can Settle at Home
So, when Canada requires a foreign issuer to register, hold liquid reserves and redeem at par before serving Canadians, will Washington read prudential regulation under Chapter 17 or a trade barrier under Chapter 19? The distinction sets the rules of engagement, and the GENIUS Act does much the same in reverse. Under section 18, the Treasury decides which foreign regimes count as comparable, and no foreign issuer reaches American customers without that blessing, or without parking reserves in the United States. If registration and reserve requirements are protectionism when Ottawa imposes them, they must be protectionism when Washington does so, too. That symmetry is Canada’s best defence, and it is why neither jurisdiction can settle the classification alone.
In fact, Washington is already running this play abroad. In Brazil, the central bank’s instant-payments system, Pix, has grown into the dominant retail rail, and the US Trade Representative has challenged it as a measure disadvantaging foreign card networks. Pix was a sovereign choice about payments infrastructure, and it is this same choice that the United States reads as a barrier to trade. The lesson here is about sequencing. The window for building a sovereign rail closes at both ends: move too late, and network effects have already handed the market to a foreign standard; move successfully, and that success becomes the very thing trade negotiation is invoked to undo. Canada might take note. Its own Real-Time Rail, first promised for 2019, is due to launch late this year, into a landscape where US dollar-denominated alternatives are already scaling.
The CUSMA review cycle is the moment to redraw the line by function rather than by form. Canada should not reopen Chapter 19 wholesale — its data flow and source code protections serve Canadian firms, too. Rather, it should pursue two things alongside the Real-Time Rail. The first is interpretive clarity: a joint understanding that instruments performing monetary functions (including payment stablecoins) fall within the financial services chapter regardless of the issuer’s corporate form, with the prudential exception applying in full. The second is a bilateral payments arrangement, using the GENIUS Act section 18’s reciprocity mechanism as the vehicle: a mutual recognition agreement of the kind securities regulators on both sides of the border already operate, pairing home-country supervision with host-country protections.
Both implementations, however, require Washington’s agreement. In contrast, the regulations under the Stablecoin Act do not. The Stablecoin Act is where Canada can settle — on its own terms and without waiting on a trade negotiation — the questions that CUSMA has left open: whether a Canadian-issued stablecoin may reference a foreign currency and on what conditions; what a substantially similar foreign regime must demonstrate before the Bank of Canada stands down in its favour; and how the federal perimeter meets provincial rules at the seams.
A coherent domestic framework is not merely a fallback, but rather the precondition for entering that conversation with something to offer, since comparability is only credible when there is a settled regime to compare. At the same time, Canada must preserve room to nurture other Canadian-dollar alternatives, including its own CBDC, so that the choice architecture in Canadian payments is not set in Washington.
If a stablecoin is a digital product, trade law governs, and finance ministries watch from the sidelines. If it is a financial service, the prudential carve-out and the supervisors it empowers come back into play. If it is money, central banks keep their mandate. And if it is all three at once — the likeliest answer — it will be governed by whichever regime asserts itself first and more effectively, while the others manage the consequences.
