DoorDash launches a text-to-order AI agent in Apple Messages

DoorDash announced on Wednesday that it is launching a text-to-order AI agent that lets users place food orders through Apple Messages. The company is opening a waitlist for users in the U.S. to try the new feature.

The launch puts a concrete consumer product behind the industry-wide push toward personal AI agents — software that completes tasks on a user's behalf without requiring them to open an app or navigate a website. It also sharpens DoorDash's competition with Uber Eats and Grubhub, its two main rivals in U.S. food delivery.

What can the DoorDash AI agent actually do?

The agent accepts natural-language prompts sent as ordinary text messages. DoorDash's flagship example is the instruction "order my usual." According to the company, the agent understands that this refers to the user's Friday night order — a recurring purchase pattern it can recognize and act on without further prompting.

Beyond repeat orders, the agent supports several other interactions:

• Specific dish requests. A user can ask for a particular item, and the agent will search local restaurants that carry it.
• Local recommendations. Users can ask for a suggestion nearby, and the agent will propose options and build a suggested cart based on the prompt.
• Photos of recommendations. The agent can text users pictures of the food it recommends, letting them see the suggestion before committing.
• Group orders. When ordering for multiple people, the agent can handle mixed dietary preferences and different quantities within a single order, DoorDash says.

That last capability matters for delivery platforms. Group ordering has long been a friction point in food delivery apps, where one person typically assembles a cart on behalf of several people with different preferences. If the agent can reliably consolidate those requests from a text thread, it removes a step that currently forces users back into the app — the exact workflow DoorDash says it wants to eliminate.

Why is DoorDash doing this now?

The company is explicit about the competitive stakes. By launching an AI agent for food ordering, DoorDash says it is looking to gain an edge over rivals Uber Eats and Grubhub. All three platforms compete on selection, delivery speed, and fees; a conversational ordering channel is a differentiator none of the rivals has matched at scale with a comparable text-based agent tied to Apple Messages.

The timing also aligns with a broader industry shift. The launch comes amid what DoorDash characterizes as a growing push to adopt personal AI agents that handle tasks without requiring users to navigate apps or websites themselves. Major technology companies are racing to position AI assistants as the front door to commerce, replacing typed searches and tap-through menus with conversational requests. Food delivery is a natural test case: orders are frequent, repetitive, and time-sensitive, which makes them well suited to an agent that can learn a household's habits.

There is also a retention logic at work. An agent that knows a user's "usual" Friday night order creates a switching cost — the accumulated preference data lives with DoorDash, not with the user's memory of a restaurant menu. The deeper the agent's knowledge of a household's ordering patterns, the harder it becomes for a competitor to displace.

The choice of Apple Messages as the interface is itself a strategic detail. Rather than requiring users to download a new app or learn a new chatbot, DoorDash meets them inside a messaging app they already open daily. That lowers the adoption barrier considerably compared with a standalone assistant.

How do you get access?

DoorDash is opening a waitlist for users in the United States to try the new feature. The company has not announced a full public release date, meaning early access will roll out gradually to waitlist sign-ups.

What else did DoorDash announce?

Alongside the AI agent, DoorDash said it will begin testing its delivery drones with select restaurants in Northern California. The drone tests signal that the company is investing on two fronts at once: software that changes how orders are placed, and hardware that changes how orders are fulfilled.

Taken together, the two announcements sketch DoorDash's direction — automating both ends of the delivery pipeline. If the text-based agent shortens the path from craving to confirmed order, and drones eventually shorten the path from restaurant to doorstep, DoorDash is positioning itself to cut friction and cost out of every stage of the transaction.

The open question is reliability. Conversational agents live or die on whether they interpret requests correctly on the first try, and a misread "order my usual" that delivers the wrong food to a group with mixed dietary needs would erode trust quickly. The waitlist period will function as a live test of whether the agent's understanding of routine orders holds up across real households — and whether texting a chatbot for dinner becomes a habit or a novelty.