Micro case studies: The feud over interest rate caps and the murky future of agentic commerce

1) Interest rate caps are great for fintechs, and the product strategy shows it

Trump’s proposed one-year cap on credit card interest rates sent shockwaves through financial markets last week, triggering immediate price declines across major banks and card issuer stocks. While the proposal hasn’t been enacted, it’s already reshaping strategic calculations across the financial sector – and fintech leaders are seizing the moment.

The back story

Credit cards are not enjoying the greatest start to the new year. Last week, Trump proposed a one-year cap on credit card rates, stating that higher rates are negatively impacting consumers.

The announcement has already led to a drop in stocks for major banks and credit card providers. 

Trump’s announcement on rate caps is still just that, an announcement, but if enacted could quickly impact their profitability. 

Meanwhile, fintech CEOs have been quick to chime in on the subject – supporting the President’s move to hem in interest rates: 

“If this is enacted—and that’s a big if, though part of me hopes it is—we would likely see a significant contraction in industry credit card lending. Credit card issuers simply won’t be able to sustain profitability at a 10% rate cap,” said SoFi’s CEO, Anthony Noto, on X. 

Similarly, Klarna’s CEO Sebastian Siemiatkowski said in a recent podcast: “In my opinion, [it’s] a very thoughtful and good suggestion from Trump to cut it to 10%. It would have returned maybe $20 billion of that back to US consumers. It’s not uncommon. We’ve seen interest rate regulation in Europe work pretty well.”

On face value, the interest rate caps seem to combat mounting credit card debt, however, it may end up negatively impacting credit availability, specifically for those who need it most: SMBs and less affluent consumers. 

For fintech CEOs, this is a good thing: In a future with interest rate caps, consumers that require access to liquidity but can’t qualify for credit at banks will turn towards fintech products. Affirm and Bilt are waiting to benefit and already making moves. 

The masterplan

In tandem with Trump’s announcement on social media, Bilt came out with three new cards that have a 10% cap on interest rates for one year. The premium Palladium option charges $495 yearly and provides credits worth $400 for hotels plus $200 in points usable at partner merchants. The mid-tier Obsidian card costs $95 annually and includes bonus rewards for dining and groceries. The entry-level card is free and gives cash back along with points on select purchases.

While Bilt chose to align its new credit cards’ announcement with Trump’s statements, Affirm is dipping its toes in a new territory through its partnership with fintech Esusu. The BNPL company will soon allow renters to pay their rents in installments. The offering is yet to be announced formally.

At the same time Affirm has also announced that it will be adding additional capabilities to its underwriting platform, adding data such as account balances and cash flow trends. 

By entering rental payments and improving its underwriting platform, Affirm is making an active effort to be a better underwriter as well as a more widely available source of credit – just as banks foresee challenges in the wake of the Trump interest cap announcement. 

The strategy is deceptively simple: millions of Americans don’t have timely rent payments reported to credit agencies, missing out on a chance to build credit history. Affirm can open doors for this functionality to its already wide user base, while fueling the sophistication of underwriting capabilities through rent data and cashflow insights. All of this will allow the company to capture a bigger chunk of consumer spend, just as consumers are pushed to seek alternative credit sources in the wake of interest rate caps.

2) The future of shopping is agentic… or not?

We have all heard the buzz about how agentic commerce stands to restructure shopping and commerce entirely. However, moves by the biggest players show that the road to this new future is going to be a rocky one. 

The back story

Last year, Amazon sued Perplexity AI over the firm’s AI shopping functionality, stating that Perplexity’s AI agent automates order placement for users, while disguising its activity as human actions. According to Amazon, these actions pose a security threat to consumer data and Amazon’s own user experience, which has been optimized for human users. 

“Rather than be transparent, Perplexity has purposely configured its CometAI software to not identify the Comet AI agent’s activities in the Amazon Store,” Amazon stated in the lawsuit. 

Perplexity on the other hand is calling the lawsuit by Amazon, bullying. The company posted the following on its website:

“Amazon wants to block you from using your own AI assistant to shop on their platform. Here’s what they’re trying to prevent: You ask your Comet Assistant to find and purchase something on Amazon. If you’re logged in to Amazon (credentials in Comet are stored securely only in your device, never on Perplexity’s servers), the Comet Assistant quickly finds and purchases the item for you, saving you time for more important tasks. Or, you can ask it to compare options and purchase the best one for your needs. Comet users love this experience.”

Amazon stated in a reponse that it is less worried about loss of advertising share and more concerned that users will miss out on options to find cheaper products and delivery options, which ultimately will impact its reputation. 

The plot thickens

It is worth noting that amidst this clash with Perplexity, Amazon is facing its own backlash. Amazon’s Shop Direct functionality allows customers to peruse items from websites other than Amazon, and some of these items have a “buy for me” feature that enables an AI agent to purchase the item on the consumer’s behalf.

It’s a classic case of the Amazonian pot calling the kettle black.  

In some cases the AI agent has placed orders for items that were never listed or were out of stock. While Amazon states that it swiftly unlists any business owners that choose to opt out, many shopowners claim that their storefronts were made part of the “buy for me” feature without ever opting in. 

Behind these lawsuits and disputes over which AI agent will rule where is a deeper realization nobody is ready to acknowledge. It may be innovation-forward to say that the tech you have under development will reshape buying and selling goods, but it is definitely uncool to admit that we have no idea what the guardrails will be. 

Perplexity is not incorrect in stating that Amazon has some serious leverage to throw around in lawsuits. Also,when agents not sanctioned by the company encroach on the shopping experience, the ecommerce giant stands to lose a major chunk of its advertising revenue. 

Similarly, Amazon isn’t wrong that unsanctioned agentic activity may put its system, UX, and users at risk.

But here is the rub: With AI agents mediating purchases on behalf of consumers, firms stand to lose relationships. All that theory about making your storefront memorable and your brand recognizable is reduced dramatically when a non-human agent is parsing your website for data and the end-consumer may or may not realize which merchant they purchased from. 

So when we say that Agentic AI will change commerce, what we mean is that it will change who owns the customer, and for brands, the answer is dark: it will be the AI agents.