The AI Underwriting Rebound: Fintech’s Resurgence in 2026

I have always believed that the most fascinating turns in the tech industry happen long after the hype cycle has peaked. Over the last couple of years, the collective tech world was mesmerized by generative AI. But if you look past the chatbots and focus on the plumbing of consumer finance, you'll see a quieter revolution taking place: the algorithm-driven overhaul of how ordinary people access credit.

Back in the early 2020s, Buy Now, Pay Later (BNPL) services and AI lending platforms were the darlings of a zero-interest-rate world. Then came the brutal valuation reset as the macro environment tightened. The truth, as usual, is more complicated than a simple boom-and-bust narrative. As we navigate the summer of 2026, we are witnessing a remarkable resurgence of the sector's survivors. And this time, they aren't selling utopian visions of disrupting banks; they are bringing hard underwriting data and massive institutional commitments to the table.

Take Affirm Holdings (AFRM.US) as a prime example. I'm told that the US-based BNPL giant, founded by PayPal alum Max Levchin, has seen its stock rally more than 50% over the past three months. This isn't just speculative fervor. In its fiscal third quarter ended March 2026, Affirm posted a 35% year-over-year surge in Gross Merchandise Volume (GMV), marking its 10th consecutive quarter of growth exceeding 30%. The platform now boasts nearly 26.8 million active consumers, with user engagement deepening to 6.7 transactions per active user.

This matters because, against the backdrop of a still-uncertain 2026 economy, Affirm is managing aggressive expansion while keeping its financial house in order. Revenue jumped 33%, and adjusted operating profit soared nearly 62% to USD 281 million. Through June and July 2026, they rolled out payment integrations with brands like Backcountry, Bed Bath & Beyond, and Overstock. When consumers are still reliably financing outdoor gear and home goods on biweekly plans—and the company explicitly notes no deterioration in consumer repayment capacity—Wall Street is naturally going to reward that underwriting discipline.

But that's only half the story. On the other end of the credit spectrum, we have Upstart Holdings (UPST.US), an AI-native lending platform determined to replace traditional credit scores with non-traditional variables like education and employment history. Their vision of an always-on, all-encompassing credit shop is also catching a powerful second wind.

Analysts have grown notably bullish. Following a management meeting in early July 2026, Needham reiterated a Buy rating on Upstart, arguing that its focus on subprime personal loans and AI-driven underwriting is the right recipe to get the stock back on track. Goldman Sachs and Bank of America have similarly raised their price targets. This restored confidence—reflected in its recent stock price recovery—is anchored in solid fundamentals: Q1 volume skyrocketed 61% year-over-year, driving a 44% increase in revenue. As they prepare to release their Q2 2026 earnings in early August, the market is betting heavily on the resilience of their proprietary models.

And yet... we shouldn't ignore the vital mechanism quietly powering both platforms: forward flow agreements. This is why market exuberance always warrants a closer look. These platforms act as aggregators, standing between consumers and institutional capital. In late June 2026, Upstart secured an agreement with Neuberger Specialty Finance to invest up to USD 600 million in consumer loans originated on its platform. Affirm went even bigger, renewing a colossal partnership with CPP Investments that commits USD 1.7 billion—with an option to expand to USD 2.2 billion—expected to support a staggering USD 8 billion in loan volume.

My view is that the 2026 performance of Affirm and Upstart proves that non-traditional data and machine learning can indeed survive a full credit cycle. They are carving out structural advantages over legacy banks. But ultimately, this remains a highly leveraged game dependent on the cost of capital. If these institutional whales ever pull their funding, the AI credit revolution could hit a brick wall. Until then, the models are holding the line. Good luck to anyone betting against them.

This article does not constitute investment advice.