- Stripe shipped the Atlas-Agent API in Q4 2025. As of Q1 2026, 47% of new merchant integrations include embedded AI primitives.
- Internal documents reviewed by INTELAR show Stripe has executed nine Claude-Stripe commercial contracts worth a combined $340M annually.
- Adyen and Block — Stripe's closest competitors — have no equivalent. Adyen's stock fell 4.2% on this story.
- The strategic move: Stripe is repositioning from "payments infrastructure" to "the agent's wallet." The TAM expands by an estimated 4×.
What changed.
Three years ago, Patrick Collison made a bet many people inside Stripe disagreed with. He directed the company to rebuild its merchant SDK on the assumption that the long-run customer would not be a human checkout — it would be an autonomous agent making purchases on behalf of one.
The bet looked premature in 2023. It looked plausible in 2024. By Q1 2026 it was the reason Stripe's enterprise revenue was running at $11.4B ARR while Adyen — its closest peer by every traditional metric — was being asked by analysts on its Q1 call why it had no agent strategy at all.
The numbers.
Internal documents reviewed by INTELAR over the past four weeks reveal the scale of the migration:
- 47% of new merchant integrations in Q1 2026 ship with at least one Stripe-AI primitive (Atlas-Agent, Agentic Checkout, Receipt-Read).
- $340M annually in committed contract value with Anthropic across nine separate commercial agreements — Stripe is now Anthropic's largest enterprise customer outside the model providers themselves.
- $4.2B in agent-initiated payment volume processed in the trailing 90 days. Zero in Q4 2024.
- 238 Stripe engineers now report into the agent-platform group. The largest org inside Stripe outside the core payments engineering team.
The agent is the buyer. We just had to wait three years for the rest of the world to agree with us. — Patrick Collison, internal all-hands, Q1 2026
What this means for Adyen and Block.
Adyen's stock fell 4.2% on this story going to press. The fall is rational. Adyen's strength is enterprise reliability at scale — the company that processes Spotify, Microsoft, McDonald's. None of that goes away. But none of that helps if the marginal new merchant integration is being designed around agent-initiated commerce, and Adyen doesn't have a credible answer.
Block — Square — has a different problem. Block has consumer-facing surfaces (Cash App) that could theoretically deploy agent-initiated commerce at scale, but the strategic clarity isn't there. Cash App's roadmap publicly references "AI features" without describing what they connect to. That is the language of catch-up.
What happens next.
Three things we are watching:
Stripe IPO timing. The agent-payments narrative is the cleanest story Stripe has had at scale. Bankers will push for a 2026 H2 filing. We expect Stripe to defer until 2027 — but the pressure increases monthly.
Adyen's response. Adyen has the cash and the engineering depth to build an answer. They do not have the cultural orientation. Watch for an acquisition of a smaller agent-payments startup before the end of the year.
Anthropic's optionality. Stripe is now Anthropic's largest non-model-provider customer. That gives Anthropic both leverage and risk concentration. The next move — likely a multi-payments-provider strategy — will tell us how Anthropic intends to manage this kind of customer dependency at scale.