Why interest-free BNPL may win when bots do the checkout

If you build or sell through automated checkout, a comment from Affirm's chief executive is worth sitting with. PYMNTS reports that Max Levchin expects AI shopping agents to push more buyers toward interest-free installment plans, because an agent comparing payment options will favor the one that costs the shopper nothing extra. That sounds like good news for consumers. For the lender saying it, it is a more complicated bet.

The money is in the loans agents would avoid

Affirm's business splits into two kinds of financing. Interest-free plans, where a merchant pays to offer zero-rate installments as a sales incentive, were 28% of the company's gross merchandise volume in the September quarter. Interest-bearing loans, where the shopper pays the financing cost, were the other 72%. That second bucket is where consumer-paid revenue concentrates.

A human shopper picks a payment plan on impulse, emotion, or whatever option is pre-selected at checkout. An agent does not. Give it a rule to minimize total cost and it will pick the zero-interest plan every time one is offered, and decline financing entirely when none is. Levchin is essentially predicting that his customers' agents will steer away from his most lucrative product.

The cost just moves, it doesn't vanish

Interest-free to the shopper is not free to provide. Someone funds the spread, and in these plans it is the merchant, through a fee baked into accepting the installment option. So agentic shopping does not remove the financing cost from the system. It relocates it from consumers to retailers, and it does so at scale, automatically, on every transaction where an agent finds a zero-rate path.

That is the part the headline framing skips. If agents reliably route volume to merchant-subsidized plans, merchants will notice their blended cost of offering BNPL rising. Some will negotiate harder, cap which products qualify, or pull the zero-interest option on thin-margin items. The equilibrium Levchin describes only holds if merchants keep paying to be the agent's default choice.

What this means if you're in the checkout flow

For retailers, the lesson is to model BNPL as a variable cost that agents will maximize, not a conversion perk used sparingly. For lenders, the pressure is to find revenue that survives a cost-minimizing buyer, which points toward merchant-side fees, agent integration deals, or interchange rather than consumer interest. For the agent builders themselves, there is leverage here: whoever controls the comparison logic effectively decides which lender and which plan wins the sale.

There are real limits to how fast this arrives. Agentic checkout is still early, most purchases are still made by people clicking buttons, and a single quarter's 28% figure tells us nothing yet about a trend. Levchin is describing a direction, not a result, and he has an obvious interest in sounding unbothered by a shift that could hit his best margins. The honest read is that he has spotted a pressure on his own model and is getting ahead of the story. Whether interest-free share actually climbs as agents spread is the number to watch.