Your message has been sent.
We’ll process your request and contact you back as soon as possible.
The form has been successfully submitted.
Please find further information in your mailbox.
Select language
Sep 23, 2026
6 min read

On September 10, 2026, Canada’s Office of the Superintendent of Financial Institutions (OSFI) said that tokenized deposits are not legally distinct from traditional deposits. The technology used to build or deliver a product doesn’t decide its legal nature. What matters is what the product is.
Most of the coverage read this as a crypto story. I would read it as a story about technology neutrality, and as the clearest confirmation so far of an argument I made back in March in Banks Will Tokenize First: banks will tokenize financial instruments at scale before startups do, because they don’t need new permission to do it.
The substance is narrow and worth reading carefully.
A tokenized deposit stays a deposit offered by the institution. The bank keeps responsibility for the underlying obligation. Distributed ledger technology changes how the claim is represented and transferred, not what it is under the existing framework. That is the whole ruling.
Now the part that got much less attention. Institutions remain responsible for the existing legal, technology, cyber and third-party risk requirements, and OSFI expects banks to consult their lead supervisors before launching a novel product. So the legal question is settled, and the operational one is wide open. Those are two different problems, and the second one is the expensive one.
I would frame it this way for any bank reading it: you no longer have to argue that your product is legal. You still have to prove that you can run it.
A tokenized deposit remains a deposit under the existing banking framework, even when the underlying technology changes.
The argument I made in March was about structure. Whether regulators would turn friendly never entered into it.
Tokenization at scale needs a set of preconditions: a license, compliance infrastructure, custody capability, institutional relationships. A startup builds those from zero, spending years and millions before the first token moves. For a bank, they are existing overhead. The advantage that creates is permission, and no amount of engineering substitutes for it.
Permission also works asymmetrically. When a bank tokenizes a deposit, it exercises a right it already holds. When a startup tries to issue a tokenized investment product, it asks a regulator for a new right, inside a framework that may simply refuse to grant it. I have worked through more than a hundred tokenization concepts, and the failure pattern is consistent: roughly 90% die at the regulatory stage, long before engineering or the business model ever gets tested. I broke that failure taxonomy down in What Kills Most Tokenization Projects.
OSFI just wrote that asymmetry into supervisory language. A deposit is a deposit regardless of the rail. That sentence is worth more to a Canadian bank than any amount of crypto legislation, because it removes the only question that was genuinely open.
This is the part I would stress with clients, because one regulator can look like a local event.
Look at what banks were already doing while waiting for nobody’s permission. JPMorgan runs roughly $7B daily in tokenized deposits through Kinexys, between its own clients. BNY launched Digital Cash in January 2026 with six institutional participants including ICE, Citadel Securities and Circle. The UK Regulated Liability Network moved from experimentation with 11 entities to a live pilot with 7 banks running through mid-2026. In almost none of these cases did anyone acquire a new license. A tokenized deposit is a deposit under banking law, a tokenized fund is a fund under securities law, and the entire compliance apparatus carries over unchanged.
The US drew the same line from the opposite direction. The GENIUS Act of July 2025 created a federal stablecoin framework with one decisive provision: stablecoins cannot pay interest, because they are payment instruments. Tokenized deposits sit in the other box. They earn yield; they live under banking law. Different jurisdiction, different mechanism, identical logic: the regulator classifies by economic and legal substance, and the rail is irrelevant.
So OSFI stated plainly what was already implicit in how every serious supervisor has been treating this.
Here is where I would push back on the optimism a bit.
Legal clarity removes an argument. Building the product is a separate job, and it is the one OSFI actually described: cyber risk, third-party risk, technology risk, and a supervisory conversation before launch. In practice, that means key management and custody design, the reconciliation layer between the on-chain representation and the core banking ledger, incident response for a system where settlement is final, and a vendor chain that a supervisor will accept. All of that is exactly as hard on September 11 as it was on September 9.
The bigger structural limit is untouched. Legal equivalence inside one institution does nothing for fungibility between institutions. A JPMorgan tokenized deposit and a BNY tokenized deposit are both US dollar demand deposits at regulated banks, and they remain non-interchangeable. Each bank mints its own on-chain money inside its own walls. I call this the cash island problem, and a ruling like OSFI’s makes islands easier to build while leaving the bridges exactly where they were. Canadian banks can now issue with confidence, and they will issue into separate pools.
That gap is the real opportunity, and I wrote about where it gets solved in DeFi Is the Missing Liquidity Layer for Tokenized Assets.
Tokenized deposits may be easier to issue, but interoperability is still unresolved. Each bank can build its own cash island, while the bridges between them remain missing.
We scope these from the legal structure up, not from the contract down
Your message has been sent.
We’ll process your request and contact you back as soon as possible.
This October, Innowise is heading back on the road, with a new US roadshow taking us across 13 cities from Detroit to Chicago.
If there’s one thing our previous roadshows have shown us, it’s how much better some conversations are in person. It’s a chance to catch up properly, talk through ideas without rushing, and sometimes discover opportunities that simply wouldn’t come up over email or on a call.
This time, Vasili Kovalevich, SVP of Business Development, Egor Grishenko, VP of Business Development, and Egor Chareichyk, Sales Director, will be travelling across the US throughout October. They already have meetings lined up with some of our current clients, but we’d also love to meet new people along the way.
So, if you’re based in one of these cities and have a project in mind, a technology challenge you’d like to talk through, or you’re simply curious about what working with Innowise could look like, come and say hello.
It doesn’t have to be a formal meeting or a big presentation. Sometimes a coffee and a good conversation are all it takes to see whether there’s something worth exploring together.
Going to be nearby? Let’s meet. Get in touch with our team and we’ll find a time that works.