Hello, and welcome back to Source of Truth, the new podcast miniseries from Fintech Takes, sponsored by our friends at SOLO.

This series is about information asymmetry (the enemy of financial services; especially lending). 

In Episode, we tackled truth vs. trust. And in Episode 2, we explore history vs. overconfidence, which goes something like this:

In 1995, Fannie Mae and Freddie Mac effectively mandated the use of FICO scores in mortgage underwriting. Almost overnight, credit scores went from a niche tool in bankers’ back rooms to the market’s law of the land.

Joining me to unpack that history and its consequences is Martin Kleinbard, Founder of Granual Fintech (and author of the great research report How Cashflow Data Can Diffuse the Credit Score Time Bomb, which I was proud to publish at Fintech Takes). 

Highlights include:

What began as an additive tool for underwriting quickly became a substitute for judgment (lenders skipped the W-2s and leaned entirely on the number)

The three pillars of credit: Willingness to pay (FICO’s wheelhouse), ability to pay (cash flow, income, assets), and product risk (loan terms that can themselves trigger default) – the financial crisis showed what happens when you ignore the latter two!

Second-order risk today. From BNPL quirks to payment hierarchy surprises (why personal loans sometimes get paid before mortgages), context still matters more than any one score

This episode explores what happens when we reduce everything to a single number: lenders miss nuance, consumers get misread. A credit score can predict repayment, but only context explains it.

And in the end, context is king.

Subscribe now and catch the rest of Source of Truth … we’re just getting started!

This miniseries is brought to you by SOLO.

SOLO resolves and connects customer data across silos — so teams stop rekeying the same customer info for the hundredth time and finally move forward.

Break the cycle at SOLO.one - That’s SOLO dot o-n-e.

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Follow Martin:

LinkedIn: https://www.linkedin.com/in/martin-kleinbard-6122aa1a/

We also chat about his new piece in Open Banker (on AI for consumer bankers), which you can read here: https://openbanker.beehiiv.com/p/aiforconsumerbankers

Learn more about SOLO here.

Hello, and welcome to Source of Truth, a new podcast miniseries from Fintech Takes, sponsored by our friends at SOLO.

This series is about information asymmetry (which is *the* enemy in financial services — especially in lending). 

Information asymmetry (the gap between what the customer knows and what the financial service provider knows) explains most inefficiencies.

It’s the reason why qualified applicants get declined for loans or approved for loans at rates much higher than they should have to pay. And it’s the reason why borrowers and human underwriters spend a mind-numbing number of hours locating, aggregating, and verifying information.

We’ve added APIs, alternative data, and automation. Yet, the core process of collecting and trusting information still looks much like it did decades ago.

Why is that? And what can be done to reshape that process in a more fundamental way? These are the questions that Source of Truth will explore.

In Episode 1, I’m joined by Georgina Merhom, Founder & CEO of SOLO. 

Georgina, a data scientist by training, unpacks how lending data is collected, verified, and (too rarely) reused. Highlights include:

What the old-school branch lending process got right that digital still misses

Why banks still burn through $30B each year on manual collection despite APIs

The difference between standardized inputs (useful) and standardized outputs (misleading)

This opening episode tees up the larger theme of the series: building systems that don’t just capture truth but create trust. 

Because when information can be reused and verified, lenders and borrowers stop starting from zero.

Subscribe now and catch the rest of Source of Truth.

This miniseries is brought to you by SOLO.

SOLO resolves and connects customer data across silos — so teams stop rekeying the same customer info for the hundredth time and finally move forward.

Break the cycle at SOLO.one - That’s SOLO dot o-n-e.

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Follow Georgina:

LinkedIn: https://www.linkedin.com/in/georginamerhom/

Learn more about SOLO here.

Bank Nerd Corner is back with Kiah Haslett returning … not just as co-host, but as an official member of Fintech Takes! 

That’s right, big news: Bank Nerd Corner will soon be its own podcast feed, with Kiah hosting (and Alex dropping in monthly as a guest). 

Kiah’s podcast launches this September alongside her new weekly newsletter, Fintech Takes Banking! If you’re listening to this episode, you basically asked for it (sign up at fintechtakes.com/banking/newsletter-subscription).

Now, onto Bank Nerding! 

First up,  the topic that's going to end up on my tombstone when I die: open banking. 

We dig into the CFPB’s sudden flip on open banking. JPMorgan Chase tried charging for data access, the Bureau hit pause on litigation, and now an accelerated rulemaking process is underway. Will banks get the green light to price data, or did Chase just overplay its hand? Is this the beginning of monopoly pricing in disguise?

Next, Kiah schools Alex (and the rest of us) on why crypto firms are suddenly obsessed with national trust charters (what they are, why they matter, and how they could function as narrow banks in disguise). Stablecoin reserves, custody rules, and OCC oversight are all on the table.

And finally, the Palmer Luckey-backed digital bank Erebor enters the chat, promising to be the new Silicon Valley Bank for startups, crypto, and defense companies. Their pitch: political connections will fast-track their national bank charter with the OCC. But can political connections really expedite a de novo charter without wrecking regulators’ credibility?

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Kiah:

LinkedIn: https://www.linkedin.com/in/khaslett/

Twitter: https://twitter.com/khaslett

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Welcome back to the Fintech Takes podcast. I’m Alex Johnson, joined by Dan Murphy — Founder of Sunset Park Advisors and former CFPB official who helped craft the agency’s open banking rule (finalized last October).

Our plan is simple: for listeners less steeped in the regulatory process, we’ll walk through how the rule took shape, assess where things stand today, and focus on what comes next (and what should come next, realistic or not.)

Granted, “today” remains a moving target. Just 20 minutes before we hit record, breaking news dropped that changed the open banking conversation yet again.

Highlights include:

The unusual bipartisan and cross-industry consensus (banks and fintechs alike) that pushed the rule across the finish line (and why that consensus collapsed after October 2024)

Why JPMorgan’s aggressive API fee move  rolled out while the no-fee rule was technically still in effect) may have backfired by uniting fintechs, crypto firms, merchants, and even regulators against it

The hardest unresolved questions: whether banks can charge for data access, how liability is allocated when things go wrong, how far the rule should extend beyond checking and credit cards, and what counts as legitimate secondary data use.

If you care about the future of data portability, the balance of power between banks and fintechs, or just want a front-row seat to the regulatory drama reshaping U.S. finance in real time, this is the episode you don’t want to miss.

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Dan Murphy:

LinkedIn: https://www.linkedin.com/in/danieljmurphy01/

Follow Alex Johnson: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

X: https://www.twitter.com/AlexH_Johnson

In the finale of Model Citizens: AI Compliance for Banks and Fintech Lenders (a six-part miniseries from the Fintech Takes podcast in partnership with FairPlay), we confront the biggest questions yet: 

What does financial regulation look like in the age of AI (when the very concept of governance is under strain)? And in a world where algorithms move faster than institutions can react, how do we protect consumers, ensure fairness, and keep power in check?

If you’re looking for quick answers or one-dimensional takes, this isn’t that episode. 

But if you’re looking for a candid, wide-ranging exploration of the forces reshaping financial services, law, and society, you’ve come to the right place.

With Kareem Saleh (Founder & CEO of FairPlay) as cohost, we talk deepfakes, regulatory dysfunction, and more with two experts in the trenches: PR Stark (Director of Machine Learning Research at FinRegLab) and Tom Brown (Senior Counsel at Paul Hastings).

Highlights include:

Why deepfakes could trigger the next bank run — and why most banks aren’t ready

The return of “model risk management” as the most underrated regulatory export

How agentic AI could upend overdraft revenue, shopping for a mortgage, and the regulatory calendar

Don’t miss this closing chapter of Model Citizens; a conversation as complex, candid, and full of possibility as the future itself.

This miniseries is brought to you by FairPlay.

FairPlay is an AI enablement company for financial services. They help companies build, test, optimize, validate and govern AI models. Learn more at Fairplay.ai

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Follow Kareem:

https://www.linkedin.com/in/kareemsaleh/

Follow PR:

https://www.linkedin.com/in/p-r-stark-1a9786142/

Follow Tom:

https://www.linkedin.com/in/tpbrown5/

Learn more about FairPlay here.

Welcome back to Fintech Takes. 

I’m Alex Johnson, and today we’re trying something different: a little audiobook experiment. 

I’m turning my recent deep dive, “Why Is This Happening? An Exhaustive Review of the History and Nascent Culture of the CFPB,” into a podcast episode for your listening pleasure(s) anytime, anywhere.

It’s a sweeping, inside-the-agency history of the Consumer Financial Protection Bureau (told through interviews with more than two dozen former staffers) and an investigation into why, 14 years after its founding, the CFPB is being hollowed out in full public view.

Along the way: cockroach-infested offices, chainsaw-wielding regulators, and a Mark Andreessen quote I never thought I’d have to say out loud. 

If you haven’t read it, or haven’t revisited it since it ran in June, you’ll hear the whole essay, start to finish. 

And because the Bureau has been unusually busy these last two months, I’ve added fresh updates on what the CFPB’s been up to (surprise flip-flop on open banking and intervention in the Synapse fiasco), and what those actions tell us about the future of the Bureau.

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/ 

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson
Twitter: https://www.twitter.com/AlexH_Johnson

Welcome back to Model Citizens: AI Compliance for Banks and Fintech Lenders, a six-part miniseries from the Fintech Takes podcast in partnership with FairPlay.

In some ways, this is the episode everything has been building toward. We’ve explored AI in customer acquisition, underwriting, and compliance — but we haven’t gone deep on the buzziest, most frontier concept in the mix: agentic AI.

In Episode 5, we tackle the systems that don’t just analyze or predict, but actually act. 

With Kareem Saleh (Founder & CEO of FairPlay) co-hosting, we sit down with Alex O’Rourke (fintech adviser, lawyer, and AI cofounder) and Jay Budzik (AI/ML product and tech leader; Senior VP at Fifth Third) to unpack the implications of letting autonomous agents loose in financial services.

Highlights include:

Why agentic AI breaks the “rules vs. empathy” binary that has defined human vs. machine roles (and how that opens new risks in compliance, customer service, and fraud detection)

How AI agents will reshape customer loyalty (expect your robo-butler to switch your accounts for better rates while you sleep)

Why banks might be more prepared than anyone thinks, thanks to a pre-existing culture of risk management (and why consumers remain the weakest link)

Don’t miss our deep dive into the wild frontiers of agentic AI … and what it means for the future of work, trust, and intelligent systems in finance.

Don’t forget to subscribe and catch the rest of Model Citizens; more insights to come!

This miniseries is brought to you by FairPlay.

FairPlay is an AI enablement company for financial services. They help companies build, test, optimize, validate and govern AI models. Learn more at Fairplay.ai

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Follow Kareem:

https://www.linkedin.com/in/kareemsaleh/

Follow Jay:

https://www.linkedin.com/in/jaybudzik/

Follow Alex:

https://www.linkedin.com/in/alexandra-villarreal-o-rourke-39631b28/

Learn more about FairPlay here.

Welcome back to Model Citizens: AI Compliance for Banks and Fintech Lenders, a six-part miniseries from the Fintech Takes podcast in partnership with FairPlay.

In Episode 4, we’re turning the spotlight on a part of lending that rarely gets the AI treatment: customer acquisition and marketing.

With Kareem Saleh (Founder & CEO of FairPlay) riding shotgun as cohost, we sit down with Alana Levine (Chief Revenue Officer at Fintel Connect) about the messy, high-stakes business of acquiring customers in a world where LLMs are rewriting the rules of discovery, personalization, and compliance.

Together, we explore how GenAI is shaking up everything from landing page copy to affiliate trust dynamics (and why marketers need to stop thinking like ad buyers and start thinking like systems architects).

Highlights include:

How AI lets teams go deeper on data to target customers more precisely than ever (and what happens when personalization crosses into “‘creepy”)

The rise of “no click” discovery and what it means for tracking attribution in LLMs

Why marketing compliance is an operational necessity (versus a bottleneck, and how FairPlay’s approach to fair lending fits in)

If you’re trying to scale customer acquisition with AI, avoid compliance landmines, and stay fair in the eyes of regulators, this one’s for you.

Don’t forget to subscribe and catch the rest of Model Citizens; more insights to come!

This miniseries is brought to you by FairPlay.

FairPlay is an AI enablement company for financial services. They help companies build, test, optimize, validate and govern AI models. Learn more at Fairplay.ai

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Follow Kareem:

https://www.linkedin.com/in/kareemsaleh/

Follow Alana:

https://www.linkedin.com/in/alanalevine/

Learn more about FairPlay here.

Welcome back to Not Fintech Investment Advice, where Simon Taylor and I riff about fintech companies we're absolutely not giving investment advice on (though this one may test our willpower).

We kick things off with Alix, which tackles one of the most bureaucratically brutal processes most people will ever face: settling a loved one’s estate. It’s admin + grief = chaos. Alix offers a $249 flat-fee concierge service that uses AI (plus humans) to cancel subscriptions, close accounts, and chase down deeds. It’s DTC in a space no one wants to think about until they have to (think: Chime-level brand softness meets probate-level emotional complexity).

Next up is Narrative, an AI-for-compliance startup that’s not trying to do everything (just the very specific, painful thing of parsing and resolving consumer complaints). What stood out? It’s not just trained on your written policies. It learns from how your best people make decisions. In a post-CFPB, state-by-state enforcement era, that nuance might be the difference between surviving a compliance audit … or hiring 300 more people to do what one model can.

Then there’s Ogment AI, which wants to be Shopify for agentic commerce. It builds MCP servers (think: APIs for LLMs) that let merchants make their products shoppable in ChatGPT, Claude, and co. But the big question isn’t tech; it’s trust. Can LLMs represent your brand voice in a way that doesn’t reduce you to “cheap and ships fast”? TBD, but Ogment is skating where the puck might go.

Finally, there’s SOLO, which is kind of like a new school credit bureau. One that’s trying to standardize, store, and reuse the messy contextual data that lives outside traditional credit files. Plus, it flips the economics: lenders get paid when others reuse their verified data. It’s a trust layer disguised as underwriting tech, and its success may hinge more on old-school, squishy human partnerships than the tech.

Plus, manifestations: We want the Timothée Chalamet of fintech; the operators who give a damn about striving to be the best at their craft. Often, the most profitable companies started that way and the monies followed as a byproduct of obsession with doing it right. Now that’s worth spotlighting.

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Simon:

LinkedIn: https://www.linkedin.com/in/sytaylor/

Substack: https://sytaylor.substack.com

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Companies featured:

https://www.meetalix.com/

https://thenarrative.dev/

https://www.ogment.ai/

https://solo.one/

Welcome back to Model Citizens: AI Compliance for Banks and Fintech Lenders, a six-part miniseries from the Fintech Takes podcast in partnership with FairPlay. 

With FairPlay’s Kareem Saleh (Founder & CEO) at my side, we unpack: how can banks and fintechs build fair, compliant lending systems in a time of regulatory uncertainty?

In episode 3, we’re digging into compliance in the age of AI; an area that isn’t always top of mind for builders in fintech, but absolutely should be.

Both of our guests today —  Chelsea Keegan (Chief Compliance Officer at Flex) and Eli Corbett (VP, Deputy General Counsel at Affirm) —  have built compliance functions inside high-growth fintech companies, either in their current roles or in the past. 

So we’re taking this opportunity to learn from their experience (what they’ve seen, what they’ve learned, and how they’ve approached compliance).

Together, we unpack the new realities of legal risk in fintech … and why compliance can’t just be a legal function anymore

Highlights include:

How you can't compliance your way out of a bad product (why product, engineering, and legal need to be in the room together, especially when building with AI)

Regulators are watching your blog posts. Your marketing language might be Exhibit A in a future consent order

Why nimble tech, cultural fluency, and strong governance matter more than legacy checklists or canned audits

Why a fintech’s worst-case scenario isn’t the same as a bank’s (and both sides need to plan accordingly)

Don’t forget to subscribe and catch the rest of Model Citizens; we’re just getting started!

This miniseries is brought to you by FairPlay.

FairPlay is an AI enablement company for financial services. They help companies build, test, optimize, validate and govern AI models. Learn more at Fairplay.ai

Sign up for Alex’s Fintech Takes newsletter for the latest insightful analysis on fintech trends, along with a heaping pile of pop culture references and copious footnotes. Every Monday and Thursday: https://workweek.com/brand/fintech-takes/

And for more exclusive insider content, don’t forget to check out my YouTube page.

Follow Alex: 

YouTube: https://www.youtube.com/channel/UCJgfH47QEwbQmkQlz1V9rQA/videos

LinkedIn: https://www.linkedin.com/in/alexhjohnson

Twitter: https://www.twitter.com/AlexH_Johnson

Follow Kareem:

https://www.linkedin.com/in/kareemsaleh/

Follow Chelsea:

LinkedIn: https://www.linkedin.com/in/chelsea-keegan/

Follow Eli:
LinkedIn: https://www.linkedin.com/in/elizabeth-eli-corbett-1b0177130/