Despite widespread speculation that artificial intelligence agents are on the verge of reshaping how consumers shop and order goods online, the actual numbers tell a more modest story - at least when it comes to food and grocery delivery. DoorDash, one of the largest delivery platforms in the United States, disclosed during its most recent quarterly earnings call that third-party AI agents have collectively placed only a few hundred orders through its platform to date. To put that figure in perspective, the company reported in its 2025 annual data that it handles an average of roughly 3 million deliveries every single day.

The disclosure came in response to analyst questions during DoorDash's third-quarter earnings call, where company CEO Tony Xu addressed the growing conversation around agentic commerce with a measured and candid assessment.

"I would say that this is something that is still a very dynamic area. And my take on all things technology is you never want to fight gravity. But at the same time, you don't want to be delusional. What I mean by that is if you look at where AI agents are today, we're seeing maybe several hundred orders in total from some of these third-party agents."

The third-party agents Xu referenced include platforms such as Muse, Instinct, and Fo, among others, all of which have positioned themselves as tools capable of handling everyday tasks - including placing food and product orders - on behalf of users. While these services represent a growing category of autonomous software, the actual volume of transactions they are generating on DoorDash remains negligible compared to the platform's overall order flow.

Users Still Prefer the App, But DoorDash Sees Future Potential

Xu confirmed that the overwhelming majority of DoorDash customers continue to place their orders directly through the company's own mobile application. The traditional app-based experience remains the dominant channel, and there is no indication that agentic ordering is displacing it in any meaningful way at this stage.

That said, DoorDash is not dismissing the concept. Xu indicated that the company is actively working to understand how agentic use cases might develop over time, and that it is interested in cultivating productive relationships with the developers building these third-party agents. He also pointed to a potential commercial opportunity down the line, suggesting that advertising within AI agent interfaces could become a viable revenue stream for DoorDash as the space matures.

DoorDash's Own AI Efforts Are Already Underway

While the company is cautious about overstating the impact of external AI agents, DoorDash has been quietly building out its own suite of AI-powered tools. In June, the company launched a chatbot called Ask DoorDash, which allows users to interact with the platform through natural language prompts and even photographs to place orders. This was followed in July by the introduction of a command-line interface agent aimed at more technically oriented users. Then in September, DoorDash rolled out a text-to-order agent, further expanding its range of conversational ordering options.

Beyond consumer-facing tools, DoorDash also released a Model Context Protocol (MCP) integration designed to enable corporate clients to place orders through external chatbots and AI platforms. This move signals that the company is thinking carefully about how enterprise customers might interact with its services through third-party software ecosystems, even if the consumer side of agentic ordering remains nascent.

A Divided Industry Response to AI Agent Access

The broader retail and commerce landscape is still working through how to respond to the rise of personal AI agents. These automated tools, which can browse websites, compare prices, and complete purchases without direct human input at each step, present both an opportunity and a challenge for established platforms.

Major retailers are taking notably different stances. Amazon has moved to restrict personal AI agents from accessing its platform, while Walmart has taken the opposite approach, actively opening its doors to agent-based interactions in an apparent bid to attract customers who might otherwise shop with competitors. The contrast illustrates just how unsettled the industry consensus remains on whether AI agents represent a threat to be managed or a channel to be embraced.

For now, the data from DoorDash suggests that the agent-driven commerce revolution, while generating considerable discussion in technology and business circles, has yet to translate into meaningful transaction volume on one of the country's largest delivery platforms. Whether that changes as the underlying technology improves and consumer adoption grows remains to be seen, but DoorDash appears to be positioning itself to participate in that future rather than resist it.