Technology

You Can Now Text OpenAI's Dot AI Agent, and It Can Text You First

Martin HollowayPublished 2w ago3 min readBased on 10 sources
Reading level
You Can Now Text OpenAI's Dot AI Agent, and It Can Text You First
source:openai.com

OpenAI has started a preview that lets US Pro subscribers text their Dot agent. On an iPhone, the texting runs through iMessage, Apple's built-in messaging service. On an Android phone, it runs through RCS, the newer texting standard that replaces older SMS messages.

Users can send texts, photos and voice memos to a Dot. A Dot can also contact its user first, sending text updates about tasks it has been given.

Dot was introduced on September 29, 2026, alongside a new $500 paid tier. OpenAI CEO Sam Altman described the agent as remarkably capable and always on.

An AI agent is a program that carries out tasks for you, rather than only answering questions. OpenAI calls Dots proactive assistants. They keep working on complex projects and everyday jobs, and they can research, analyze data and prepare documents. Bloomberg reported in October that Dots are aimed at OpenAI users who already use its products at work for writing code and producing research.

Setup happens inside ChatGPT. You open the Dot's profile and select Texting. The feature needs iMessage turned on for an iPhone, or RCS for an Android phone. OpenAI uses Linq, a messaging company, to provide the texting behind it.

Text is one option among several. OpenAI says you can message or call your Dot inside ChatGPT, and you can type while talking to it. Bloomberg also reported at launch that you can reach a Dot by sending a message on Slack, the workplace chat app.

Memory connects all of these. Anything a Dot saves to its memory is shared across ChatGPT and text, so context carries from one conversation to the next. If you set up a task in the app and then follow up by text, you are talking to the same agent, with the same stored information.

OpenAI also offers a separate Messages integration, and it works differently. Its ChatGPT release notes describe an Apple Messages plugin in the ChatGPT desktop app. The plugin can read and search iMessage, SMS and RCS conversations, and it can prepare or send messages through Messages. The plugin documentation says that in Codex and ChatGPT Work, the integration can read and search those chats on a Mac and send messages on your behalf. It works from your own message history on the Mac. A Dot is different, because it keeps its own thread with you.

In my view, the notable change is the interface. An agent that lives in a message thread you already use needs no new app, no dashboard and no new habit. Giving it a job then looks like sending a message to a coworker, which is a lower bar than opening a control panel.

The open questions are in the details. Technical readers will check how a message the agent starts is told apart from one you start, what the agent is allowed to send for you, and how its memory is kept and shown when a conversation moves across several channels. Those questions will matter more as the feature moves beyond the Pro preview.

The broader context here will be familiar to anyone who has followed messaging apps. Each time a service moved into text-style threads, from early chat programs to phone push notifications, it made life easier and also opened new routes for alerts, consent and mistakes. Dots add a new wrinkle, because the agent can contact you on its own about work you gave it. If the feature works as described, it could make a long-running agent feel like a normal part of the day, rather than a tool you have to remember to open. I would watch that, along with how the controls around it develop.