PayPal and Robinhood AI agents in fintech

PayPal’s shares have staged a volatile recovery from the February plunge that sent the stock as low as $38.83, but the rebound has not erased the core challenge. Even after climbing back above its 50-day moving average in late summer, the stock has slipped back to about $54.67, just above that average, with momentum indicators losing steam. That mirrors the broader story across fintech: management teams are spending more on technology and development, while the payoff from agentic tools remains more of a strategic option than a visible earnings driver.

Robinhood is the cleaner read-through on where the category is heading. The company says it already uses machine learning and AI for customer support, fraud detection and coding efficiency, and it has now gone further with “agentic trading” that lets third-party AI agents initiate and execute transactions. That is exactly the kind of shift that could reshape brokerage economics over time. But it also brings fresh regulatory, privacy and cybersecurity risk, which means near-term monetization will likely be slower than the hype cycle suggests. The stock’s ability to hold above its 50-day moving average after a sharp selloff shows investors still want exposure, but they are no longer paying any price for the story.

The deeper investment case is that AI agents will matter most as infrastructure, not as a glossy feature. The winners are likely to be the fintechs that own the rails, the data, the compliance layer and the distribution relationship, because those are the toll roads through which autonomous commerce will flow. That makes payments processors, brokerages, fraud-stack vendors and cloud-dependent software suppliers the more durable beneficiaries than the companies simply demoing AI assistants in app screens.

There is also a valuation trap here. In a market already rattled by extreme fear readings in the broader S&P 500 signal set, investors are less willing to reward vague AI promises. When risk appetite weakens, stories need proof. For fintechs, proof means lower servicing costs, better conversion, higher engagement and measurable margin expansion. Until that arrives, AI agents may remain more useful as a defense against competitive erosion than as a clean source of revenue acceleration.

I believe the market is underestimating how long it will take for autonomous finance to show up in reported numbers, but also underestimating the scale of the eventual prize. The right way to play this theme is not to chase every AI-branded fintech. It is to own the platforms that can turn automation into scale: the payment networks, the brokerage infrastructure, the fraud and risk engines and the cloud beneficiaries that sit one step behind the user interface.

For investors, that means staying selective. The first wave of AI agents may not transform fintech earnings immediately, but they are setting up the next phase of competition — and the companies with the deepest data, lowest marginal costs and strongest compliance architecture are the ones most likely to turn today’s hype into tomorrow’s cash flow.