AI Agents Move Directly Into SMS and iMessage Threads

A growing group of tech companies and developers are testing consumer workflows directly inside standard text messaging threads instead of requiring standalone app downloads. Through conversational prompts, users can coordinate schedules, organize family logistics, and order food without opening a dedicated mobile application.

Ordering Food Through iMessage

DoorDash unveiled an AI agent at its Dash Forward 2026 event that runs inside Apple's Messages app. The tool allows users to search menus across more than 800,000 restaurant and grocery products, assemble carts, and complete orders through conversational text prompts.

The system links a user's phone number to their DoorDash profile and saved payment methods. DoorDash claims the agent can process complex requests, such as filtering menu items across mixed dietary restrictions for group orders or generating a grocery replenishment list from a photo of an open refrigerator.

DoorDash co-founder Andy Fang described the shift as a new application paradigm, arguing that managing tasks like ordering lunch or groceries should not feel like work. Chief executive Tony Xu similarly noted that consumers care about practical convenience rather than the underlying terminology.

The ordering feature is currently available as a beta waitlist for iPhone users in the United States. DoorDash has not announced rollout timelines for Android devices or non-U.S. markets.

Household Coordination via Shared Messages

Google has also expanded its Google CC AI agent into a shared household assistant. Designed to function in group messaging threads, the service supports up to six adult family members.

The household assistant monitors natural-language updates in shared chats to manage collective task lists, suggest calendar events, and organize routine family logistics without requiring individual members to enter details into separate calendar software.

Developer Infrastructure for Messaging Agents

The move to embed automated assistants across SMS, RCS, iMessage, and WhatsApp has spurred interest in specialized backend infrastructure. Developer startup Photon raised a $4.5 million seed financing round co-led by Gradient Ventures and A*.

Photon provides orchestration software that lets developers run conversational agents across iMessage, SMS, WhatsApp, and email. The startup reported over 40,000 developer sign-ups, with 98 percent of current usage on its open-source framework, alongside tenfold revenue growth over four months for its enterprise tier.

Friction and Pricing Inaccuracies

Despite claims that conversational interfaces eliminate mobile app friction, early hands-on testing highlights hurdles. Testing by Bloomberg on an early build of DoorDash's text agent found that the system returned incorrect photos and quoted prices that differed from the main DoorDash app.

DoorDash stated that public waitlist users will access an updated model designed to address those early inconsistencies. Pricing discrepancies also raise untested legal questions in foreign markets, such as potential scrutiny under the European Union's Unfair Commercial Practices Directive.

Usability remains an open debate. Critics point out that texting back and forth with an automated bot can require more keystrokes and time than simply tapping a saved order in an app. Whether text-based agentic workflows will achieve mainstream consumer adoption over conventional mobile apps remains unproven.