#

Key takeaways

• Stripe Link reportedly powered about 200 agents at the time of the interview.

• Upstarts Media said Stripe had about 70 agents signed on in a September analysis.

• The figures refer to participating agents, not users, purchases, or revenue.

• The report says agentic usage is still small, but rising.

• The detailed payment breakdown was behind subscriber access.

What the report says

Upstarts Media published an October 18 report based on an interview with Dan Hill, Stripe's head of consumer product for Link. The report focuses on payments initiated by AI assistants. Hill said Link then powered about 200 agents. The publication also said Stripe had roughly 70 agents signed on when it compiled a September analysis of agentic spending.

The report is careful about what those numbers mean. They are counts of participating agents as described by Stripe. They are not counts of users, completed purchases, or revenue. That distinction matters because the figures describe participation, not commercial output.

The publicly readable portion of the report names consumer and work-oriented agents that can make purchases through Link on users' behalf. It also says Anthropic and OpenAI use Link. In the context discussed, however, they were not yet using it for agentic commerce.

How Stripe framed the trend

Hill described agentic usage as a still-small part of Link's overall reach. He also said the trend was rising. That gives the report a clear direction, but not a large-scale conclusion. The source supports a picture of early adoption rather than broad maturity.

The report also says Link has around 300 million users overall across conventional businesses. That broad user figure should not be treated as the number of people using AI agents for checkout. The source separates general Link usage from the smaller set of agentic interactions.

What is not established

The detailed breakdown of payment volume, categories, and transaction sizes is behind the publication's subscriber access. Because of that, this article does not assert figures from that unseen analysis. It also does not present the report as an independent audit of Stripe transactions.

The account is described as an interview and a company-provided snapshot. That means the source is useful for understanding Stripe's own framing, but it does not prove broader market adoption on its own. The report also does not establish that this feature is available through a Moroccan merchant or bank.

Operational and governance considerations

The source suggests that AI-assisted checkout is still in an early phase. That makes measurement important. Companies reading the report should separate agent counts from user counts, and should avoid treating participation as proof of completed commerce.

The report also shows why access controls matter. If an AI assistant can initiate purchases on a user's behalf, the system needs clear boundaries around authorization and transaction handling. The source does not provide technical details, so any deeper operational design would be an assumption.

Morocco relevance

The source reports no Morocco-specific availability or deployment. For readers in Morocco, the only supported lesson is conditional and general: if a similar payment flow is introduced locally, it should be evaluated with the same care around authorization, reporting, and transaction clarity.

Bottom line

This report presents early evidence of AI agent payment activity inside Stripe Link. The numbers are modest, the framing is cautious, and the detailed spending data is not public in the source. The clearest takeaway is that agentic commerce is being tracked, but it is still a small part of the broader Link ecosystem.