Regulatory Roulette: Stablecoins and Open Finance in 2025

Welcome to Fintech Brainfood, the weekly deep dive into Fintech news, events, and analysis. You can subscribe by hitting the button below, and you can get in touch by hitting reply to the email (or subscribing then replying)

Plaids APIs. AI Partners. Immaculate Vibes.@ Fintech Nerdcon 🤓. Who says we can’t have consumer facing AI-first Fintech? Challenge accepted. Bring your compliance worries and Let’s build. Tickets are gonna go faster than you can say 1033…

Weekly Rant 📣

Regulatory Roulette: Stablecoins and Open Finance in 2025.

Let me paint you a picture of life as a partner bank in 2025. You're sitting across from your state’s banking examiner, explaining something that should be straightforward: implementing card tokenization for Apple Pay.

You say “tokenization”

They're terrified because they don't understand what comes next. You're talking about Apple Pay. The most vanilla, well-understood type of card tokenization that millions of people use every day. Yet it takes 30 minutes to talk the examiner off the ledge and explain what card tokenization actually is and why it's not crypto.

This is a real conversation that happened. And it crystallizes everything wrong with our current moment.

Innovation is Back At Regulators 

Federal regulators are doing everything possible to provide more regulatory clarity in tokenization and Fintech.

The SEC is in the midst of a flashy tour of the country for the Crypto Taskforce, complete with social media and video advertising. The Fed Chair, the OCC, the SEC Chair have all given pro-innovation speeches in the past few weeks. In the UK, the FCA is back pushing the boundaries with AI sandboxes, and the UAE's decade-long focus on regulatory innovation has become the #2 investment location globally for Fintech. We also have the first major new financial services law in over 15 years in the US with the passage of the GENIUS Act.

I love all of this. I think on net, it's good for consumers, good for the economy.

But Reality On The Ground Is Different

If you’re trying to build in fintech or stablecoins there’s a handful of reasons your banks might be going slower than you want:

1. They’re (rightly) concerned sins committed today could be judged in 5 or even 10 years, when there's a very different regulator or administration.

2. In the U.S, the GENIUS Act has at least an 18-month wait before rule-making is complete, creating a log jam in regulators.

3. 1033 is a case in point. Dodd-Frank was passed in 2010, and we still haven't got a fully functioning rule.

4. There's a growing gap between the innovation policy side of agencies and the examiners.

5. Many state-level regulators are stepping in where federal regulators stepped back.

The result? Banks that should be racing to capture market share in stablecoins, are at least 18 months away from being able to do something meaningful, unless they can proactively demonstrate how they've got control of this space. 

If banks can’t help stablecoin companies, then new institutions will emerge who can. Their worst nightmares come true, unless they’re able to engage in innovation. 

The brave will be those who figure out the risks. And get live anyway. Not as a compliance checkbox process. But by treating risk as the R&D focus.

Sins Committed Today Could Haunt You Tomorrow

During 2015-2020, we had a Fintech version of regulatory optimism. fintech was booming driven by the rise of mobile and Neobanks. Smaller US based banks were eager to partner. "Innovation" was the buzzword du jour.

Then reality hit.

Silicon Valley Bank collapsed after a bank run triggered by social media. Silvergate and Signature followed. Synapse, the poster child for embedded banking, imploded and left thousands of consumers unable to access their money. The bankruptcy proceedings are still sorting through that mess.

Immediately the press howled, “where are the regulators?!” In reality the FDIC and others were working weekends to try and make sense of a black swan event, and prevent a potential contagion risk. 

At the state and federal level, regulators began looking into “novel activities” and the state-level examination cycle (often occurring in 12 month increments) began paying much closer attention to anything in digital assets or bank/fintech partnerships.

And what they found in some of those exams wasn’t always pretty.

Veteran readers of Fintech Brainfood will remember “BaaS is Dead” and the sheer volume of rookie errors in reconciliation, AML and poor partnership diligence by banks who optimized for being easy to work with vs slow and diligent. 

There's always a gap between what policy says and what examiners examine. But right now, that gap has become a chasm. Regulators need rules in order to enforce them. Bank partners will have a hard time supporting fintech and stablecoin businesses until those rules are available. 

People often forget, examiners are humans (no, really). These people are often under resourced, and may have one expertise or another. They have a short window to dive deeply into a bank and its business and try and understand how its performing.

The Roulette: Your Examiner May Vary

You could get the Apple Pay examiner or the crypto-traumatized examiner, and you won't know which until they walk in the room.

Different states have more experience in certain activities, and they have agencies with their own priorities. For example, New York has a Bitlicence, is very concerned with consumer protection and tends not to have a ton of experience in capital markets and banking with the HQ of some of the worlds largest banks.

If you operate nationally, no matter what the federal rules are, New York, California and larger, well-funded state regulators have to be a major consideration. 

Vs the Kentucky State, or Illinois State regulator who in 2020 may not have deeply understood bank/fintech partnerships, and had to speed-run that to catch up as we began to see failings.

Third party risk management is hard to examine if you’ve never done it before, so what do you look for? Anything. And you keep asking. That’s why a lot of banks complained about endless examinations looking for a needle in a haystack. 

Examiners didn’t know when the next contagion could appear, so they also asked banks to “pause” any crypto-related activities when they investigated. That paused ended up being indefinite and resulted in a a defacto ban on new crypto accounts (whether that was intentional or not isn’t the point it was the outcome).

When it comes to stablecoins and the GENIUS Act, examiners don’t have rules yet, and its not entirely clear when those rules will appear.

The 15 Year Gap Between Laws Passing and Rules 

There could be multiple administrations between now and when we have any rules for GENIUS that are available to regulators.

We've been here before. In 2010, Congress passed Dodd-Frank with great a specific section called 1033 buried within its 2,300+ pages. 1033 was designed to ensure a financial services provider would “make available to a consumer information in the control or possession of the provider concerning the consumer financial product or service that the consumer obtained from the provider.”

Fifteen years later, we're still not even close to implementing it.

Section 1033, the open finance rule, was kicked around for over a decade, spanning multiple administrations (Obama, Trump #1 and Biden). The Chopra-era CFPB finally passed created and passed a rule. With howls of pain from the banking lobby and jubilation from the fintech community.

Then came the election change. The new CFPB vacated the rule. Chaos ensued.

Eight years to make a rule. Fifteen years since the law passed. And we're still not done.

GENIUS Act Rules Could Take a LONG Time to Appear

It’s entirely possible that the GENIUS Act faces the same trajectory. Congress passed a law with vague language about stablecoins being legal. Now comes years of rulemaking, lobbying, and political football. And that football is already being kicked around:

The American Bankers Association, apposes Section 16(d) of the GENIUS Act that allows Permitted Payments Stablecoin Issuers (e.g. Circle, Paxos et al), to bypass state licencing for Money Transmission (payments). 

The crypto lobby disagrees. 

Banks know this pattern. They've lived through it. They know State authority is a sore spot issue to press. And they're not making big bets on regulatory clarity arriving anytime soon.

What are the implications here?

Two things.

1. The Crypto lobby is much more organised and well funded than the Fintech lobby of 2015 to 2020. And it’s also united with the Fintech lobby and trade groups. This is at least a fair fight, but that doesn’t solve our challenge if we work in the industry as operators.

2. The existence of 1033 in law and the rise of mobile led to much more open finance adoption in the US than we see in Europe. The GENIUS Act may not have a rule, but that won’t stop innovators (nor should it).

Because while this fight happens, we still need to do business.

Act As If the Rules Already Exist

For anyone in Fintech, stablecoins or crypto, banking relationships are more fragile than they appear. Even if your current partner is supportive, examination pressure could change their risk appetite overnight.

This is why, no matter how much the crypto lobby howls about “debanking” or 1033, the reality at their partner banks is that they can’t go any faster. Nothing is going to make your roadmap go faster than you figuring out how to manage the risk with your bank partners and them to their regulators.

I spoke to one CEO recently who said something that is now much more common,

Most importantly: act as if strict rules already exist. What would comprehensive stablecoin regulation require? Build those capabilities now. You'll be ready when rules arrive, and you'll demonstrate good faith if questions come up later.

For banks. So much can be achieved by being able to walk everyone through a step-by-step set of diagrams. I know this is already being done. But in unknown territory (like stablecoins) a flow of funds diagram is worth 10,000 words.

The Market Isn’t Waiting

You should be developing your stablecoin framework right now. 

You should be figuring out how tokenized deposits work alongside stablecoins. 

Your downside risk is that all this noise goes away. The upside is you're still relevant when this trend eats domestic market share.

Innovation is back in fashion. The policy environment is the most supportive it's been in decades. But the examination reality lags behind by years, not months. Banks know this. They've been through the cycle before. They're not opening champagne because they remember how the last party ended.

That doesn't mean you stop innovating. It means you innovate thoughtfully. You build relationships with regulators. You document everything. You plan for multiple scenarios. You treat compliance as R&D and a competitive advantage for the long term.

The companies that master this balance will have enormous competitive advantages when clarity arrives. They'll have battle-tested frameworks, stronger regulator relationships, and more sophisticated risk management.

The future belongs to those who can handle uncertainty without being paralyzed by it.

The easiest way out of uncertainty is to take action. 

Learn. 

Repeat.

ST.

4 Fintech Companies 💸

1. AskRose - ChatGPT for Excel

Rose is your embedded ChatGPT colleague AI. It can create entire models, make charts and can make calculations from natural language. For example “figure out my interest on a 30 year mortgage” or “figure out our engineering COGS based on these inputs”

🧠 Isn’t it wild that the company that built Microsoft Clippy made such a bad product in co-pilot there are multiple companies building competitors? Every finance team has one or two analysts who are excel wizards. That’s now a commodity. I just wonder if this can become a business. I actually hope Microsofot and/or Google acquires something like this to make their in-product-AI better.

2. Veris - The Mirror to your stack for AI Testing

Veris helps enterprises like banks create a mirror to their existing tech stack that reproduces every tool and interaction your agent will face. This allow teams to run 1000s of tests against this mirror and then collects performance data to help with reinforcement learning (RL) the agent.

🧠The enterprise adoption of AI is just getting started, this is well timed. While start-ups are facing a COGS issue, enterprises have gone from POC to starting to buy codegen tools. Enterprises sign multi-year deals, and need much more certainty in agent performance. A tool like this could fit well in the new world of cheaper AI models, RL’d to more tightly constrained use cases.

3. Bumper - BNPL for car repairs

Bumper helps customers split the cost of car repairs and servicing into monthly, interest-free payments. To date they have more than 5,000 UK auto dealers onboarded. Users select their dealership, enter their details to the bumper app, book their appointment and then spread the cost. 

🧠 Every niche gets its own BNPL provider? What’s new here other than the go to market? There’s something valuable in serving a niche, but why could every auto dealer support Klarna or Affirm? 

4. Receive - Card-linked earned revenue access for SMBs

Receive provides businesses with access to funds before they arrive. It works by connecting bank accounts and receivables data and analyzing sales data. The card has no fees, no interest or penalties and is designed to allow businesses to spend on inventory or invest in growth.

🧠I guess they’re using the card interchange as their revenue generating model. This is clearly aimed at the smaller end of small businesses vs growth companies. It’s an interesting idea to link the card spend directly to the receivables data and line of credit. Lots of innovation happening in the credit card space lately. 

Things to know 👀

1. Rain Cards raises $58m Series B to scale stablecoin linked cards

Rain helps stablecoins wallets issue credit and debit cards. Meaning A customer can now pay with stablecoins anywhere visa is accepted. Rain raised $58m, led by Rain, which allows neobanks and fintech firms, such as Nuvei, to provide stablecoin payments.

🧠 Think of Rain as “Marqeta for stablecoin wallets.” If your customers can hold stablecoins, Rain lets you give them a Visa card so they can spend those stablecoins.

🧠 This opens new markets. Stablecoin wallets allow customers in long-tail markets to spend without high FX fees or delays. Bringing them in to the global e-commerce ecosystem.

🧠 Everyone is missing what makes Rain special. 

🧠 They give issuers instant settlement from Visa. Transaction is authorized, USDC is delivered to stablecoin wallet of issuer, 24/7× 365.

🧠 This is huge if you’re extending credit. It reduces their fraud risk and means they’re much more capital efficient if they get paid sooner. You’re not waiting to find out if the merchant was a fake website, or borrowing more to lend more. (Most use cases are debit today, because wallets are pre-funded, but the potential here is huge)

🧠 What makes Rain truly special is that they've created the first real-time settlement card infrastructure for issuers. So while it’s tempting to see a big round as just investor enthusiasm for anything stablecoins, look deeper at a new settlement paradigm.

2. Chinese Gangs suspected in $312 billion transactions the financial system.

FinCEN the Anti Money Laundering division of the US Treasury issued a warning of Chinese Money Laundering Networks (CMLNs) “suspected to be linked to as much as $312 billion in illicit activity between 2020 and 2024.” and recommended looking at suspicious activity from Chinese passport holders.

🧠 This is based on suspicious activity report fillings by banks themselves. The data comes from 137,000 fillings by institutions.

🧠 That does not mean $312bn was successfully laundered, just suspected. Although often when banks flag suspicious activity, this has happened after the transaction has cleared and settled. So I’d wager a high percentage probably did.

🧠 Still, the opacity and lack of clarity here is astonishing. The sad reality is AML officers at banks are still playing whack a mole with ancient transaction monitoring systems, and they’re often the last to get tech budget. They throw bodies at an infra problem.

🧠 We need better infra. SWIFT helps banks securely share messages among its members, JPM has a highly secure cryptography solution (IIN), but the adoption here is fragmented. and limited to their networks.

🧠 What’s the network of networks here? I think work by Note Bene, and Bluprynt here is, while early, a sign of what could be done with cryptography for all transactions regardless of network.

Good Reads 📚

The AI Short Squeeze

Everyone has seen the chart of tokens getting 10x cheaper, meaning if you charge a subscription today, in a year you’ll have 90% margins. 18 months later, Windsurf is being sold for parts, and claude code rolled back their unlimited tier. The problem? When a new model comes out, 99% of usage flips to it, and the new models are all reasoning models. Those reasoning models use 10x or 100x more tokens to say “hello there”

At $1 per deep research run, a $20/month subscription cannot even support a user making a single deep research task a day. We’re locked in a prisoners dilemma, where everyone is posting hockey stick growth charts while subsidizing power users. They don’t want to charge on usage because users hate that.

🧠You can see why GPT-5 was trying to shift you to the shittier model now? Nvidia can’t build chips fast enough, and hyperscalers can’t build data centers fast enough so there is a supply issue. Demand for compute is 100x, supply is constant. Meaning everyone in AI has a COGS problem.

Tweets of the week 🕊

That's all, folks. 👋

Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)

(1) All content and views expressed here are the authors' personal opinions and do not reflect the views of any of their employers or employees.

(2) All companies or assets mentioned by the author in which the author has a personal and/or financial interest are denoted with a *. None of the above constitutes investment advice, and you should seek independent advice before making any investment decisions.

(3) Any companies mentioned are top of mind and used for illustrative purposes only.

(4) A team of researchers has not rigorously fact-checked this. Please don't take it as gospel—strong opinions weakly held

(5) Citations may be missing, and I've done my best to cite, but I will always aim to update and correct the live version where possible. If I cited you and got the referencing wrong, please reach out