Hey, fintech fam! 💜

I got off a call last week and couldn’t stop thinking about one sentence:

Innovation needs infrastructure.

It’s a simple idea. But once you start looking for it, it’s everywhere in how AI is entering financial services right now.

It’s underneath every headline about a frontier AI company connecting to your bank account. Every new financial copilot promising personalized advice. And every conversation about what happens when an AI agent stops answering questions about your money and starts acting on your behalf.

The more I thought about it, the more I realized there’s another layer of the AI race worth paying attention to.

Today’s story starts with a conversation I had with Plaid CMO Hannah Hughes about who actually earns our trust when AI becomes the interface to our financial lives.

FTW: San Francisco is one week away, and the room keeps getting sharper. If you haven’t grabbed your ticket yet, now’s the time.

Let’s get into it. ✨

Last week, Plaid Chief Marketing Officer Hannah Hughes and I jumped on a call to prep for her upcoming fireside chat closing out the Fintech Is Femme Leadership Summit in San Francisco.

About halfway through our conversation, I realized something.

If an AI product asked me tomorrow to connect my bank account so it could analyze my finances, there’s one logo that would make me much more likely to say yes:

Plaid.

And that’s interesting because I’ve probably connected my bank accounts through Plaid across a dozen apps over the years. I couldn’t name them all without checking.

That’s kind of the point.

Plaid sits underneath the experience. The app gets the customer relationship. Plaid makes the connection possible.

It’s why a financial app I downloaded 20 minutes ago can ask for access to some of the most sensitive data I have, and I can connect an account without handing that company my banking credentials directly.

More than half of Americans with a bank account have experienced some version of that handoff.

Most probably didn’t spend much time thinking about the infrastructure underneath it.

That was a feature of the last era of fintech. I’m starting to think it could become a defining advantage in the next one.

Because AI is quickly becoming another front door to our financial lives.

People are already asking ChatGPT hundreds of millions of finance-related questions. Perplexity now lets users connect checking, savings, and loan accounts to create a live picture of their finances.

But the moment AI goes from telling me what an emergency fund is to telling me what I should do with my money, it runs into a very different problem.

It needs to know me.

And knowing me requires data + infrastructure.

A foundation model can explain how much someone should theoretically save for retirement.

It cannot tell me whether I should increase my contribution this month without understanding my income, spending, debt, cash flow, accounts, and everything else happening in my financial life.

Intelligence without context can only get you so far.

Hughes kept coming back to that one phrase during our conversation:

Innovation needs infrastructure.

I haven’t stopped thinking about it.

Plaid’s consumer-permissioned data network was built for an earlier era of fintech, when thousands of apps needed a secure way to connect consumers to their financial data.

Then AI showed up.

Suddenly, the same basic problem exists at a much larger scale.

Every AI lab building a financial copilot needs context. Every wealth platform trying to personalize advice needs data. Every agent that eventually wants to act on someone’s behalf needs some combination of identity, permission, connectivity, and trust before it can touch their financial life.

The smarter the interface gets, the more consequential the infrastructure underneath it becomes.

And Plaid hasn’t exactly been standing still.

If your mental model of Plaid is still the little screen that appears when you connect your bank account, you’re missing a meaningful part of the company it has become.

Plaid has trained models using its own transaction data, reporting a 48% improvement in income classification accuracy and a 13% improvement in transaction categorization.

Its fraud technology maps relationships across accounts and devices in real time, and the company says it can identify up to 41% more fraud at the same false-positive rate.

Then there’s the customer mix.

Plaid completed a tender offer at an $8 billion valuation, up 31% from the prior year, Bloomberg reported in February. AI companies accounted for 20% of the new customers Plaid onboarded last year. 

The market is already pricing in what Hughes was describing to me, and what I’ve been touching on in past newsletters: the infrastructure layer gets busier as AI gets more capable.

As the number of companies trying to build financial experiences expands, the number of companies that need trusted financial infrastructure expands with it.

The infrastructure race is now, and the winners (like Plaid) are joining us in San Francisco from September 29th through October 1st. Don’t miss it. 

For years, Plaid's customer base was fintech apps and financial institutions that needed a trusted way to move data. 

Now add every AI lab building a financial copilot, every wealth management tool personalizing advice at scale, every agent moving toward acting on someone's behalf directly. 

Hughes left the long-term implications for Plaid’s value as an open question when we spoke, and I will too.

What I can tell you, from a decade of watching infrastructure companies get overlooked until the moment they're indispensable, is that AI isn’t making financial infrastructure less important. It’s dramatically increasing the number of products (and people) that need it.

And that’s where the story gets more complicated.

Connecting an AI to your financial life is one problem.

Letting it act on that financial life is another.

This month, Reuters reported on problems encountered during internal testing of Meta’s new agentic technology as the company pushes AI toward tasks that can include sending emails, managing calendars, and making payments.

Meta's own security incidents are up 40% year over year as agent-related issues climb. 

Anthropic launched a version of Claude for financial advisers the same week, and I wrote earlier this month about OpenAI building a coalition of licensed data providers just to make its own financial services product trustworthy enough for a bank to use.

Everyone is moving toward the same frontier:

AI that knows enough about you to become useful, and eventually has enough permission to do something about it.

Which raises a question I think financial-services leaders are going to spend the next several years answering:

Who owns the relationship when an AI interface sits between a consumer and their financial life?

And underneath that is another question:

Who does the consumer trust enough to connect the two?

Hughes described Plaid’s potential role to me as something closer to a consumer trust seal than a traditional B2B provider.

That’s a fascinating evolution for a company whose infrastructure was intentionally designed to disappear into the background.

For years, success meant you barely noticed Plaid.

In an AI economy, knowing who is underneath the experience may become part of the reason you’re willing to use it.

There’s another reason I wanted Hannah closing the Leadership Summit.

Before Plaid, she spent years building the kinds of consumer financial products that infrastructure companies like Plaid sit underneath, including lending at Affirm and financial products at Block.

And for her, questions of financial access aren’t purely professional.

She grew up in a low-income family in Alabama. By the time she arrived at Plaid, she had already spent years thinking about what happens when a financial product reaches a consumer, what promise that product is making, and what happens when trust breaks.

Now she’s sitting on the other side of the stack.

That’s the perspective she’ll bring to FTW: San Francisco on September 30, when she closes our Leadership Summit in conversation with me.

And the timing matters.

Throughout the day, we’ll follow AI further and further into the financial system: how it changes work, what happens to expertise, how financial institutions deploy it, what infrastructure companies should own, and what happens when AI agents begin interacting with the financial system themselves.

Hannah and I get to end with the question underneath all of them:

What has to exist before we’re willing to trust AI with our financial lives?

Fintech spent the last decade making infrastructure invisible on purpose.

AI may be the thing that finally makes us look underneath.

And I have a feeling we’re going to care a lot more about what we find there.

I’ll see you in San Francisco.

Executives from companies like Grasshopper, Stripe, Candidly, Green Dot, and more will be in the room September 29th-October 1st.

We’re in a golden AI era because this is the time to learn everything you can and get ahead of your competition.

In 5 years, “I don’t know how to use AI” won’t fly. Everyone will know how to use it, and companies like Plaid will have even more case studies about how useful it is.

It’s now or never. Make it now and join us at FTW: SF.

She dropped Cyrus, if you didn’t hear. Last week was old Miley; this week is new Miley (in music and in name).

📰 Share this newsletter with a friend and start growing your network.

🔗 Connect with me on LinkedIn for daily insights on leadership.

🤝 Grow your business through content & community by partnering with me.

📣 Promote yourself to 50,000 subscribers by sponsoring this newsletter or the Humans of Fintech podcast.

🎤 Host an epic event by booking me as a speaker, moderator, or emcee.

📚 Increase your expertise by ordering your copy of my book, Fintech Feminists: Increasing Inclusion, Redefining Innovation, and Changing the Future for Women Around the World.

Thanks for spending time with me today. I’ll see you in San Francisco!

Love,

Nicole 💜