OpenAI Begins Beta Texting for Its Dots Agents

An agent you can reach by text

OpenAI began a beta rollout of texting for its dots on October 28. Dots are ChatGPT agents that can continue ongoing tasks between conversations. The new option lets eligible users reach a dot through iMessage or RCS instead of returning to its ChatGPT conversation.

Open a dot’s profile in ChatGPT and select Texting to connect it. OpenAI says a dot can also send proactive updates it considers relevant. The announcement describes a limited rollout to Pro users in the United States, so availability may depend on the account and rollout stage.

Messages share the dot’s context

A text conversation uses the same memory and plugins as the dot in ChatGPT, so the agent can continue with the context it already has. OpenAI says texting does not give the dot or OpenAI access to a user’s other text conversations, and that texting works one-on-one rather than in group chats.

The feature relies on Linq to send and receive messages. OpenAI’s help page says Linq processes the messages on their way to OpenAI, which means the conversation is not end-to-end encrypted between the user and the dot, even over iMessage or RCS. Linq says it does not retain message contents after delivery, while OpenAI stores the texts as part of the dot’s ChatGPT conversation. These details are from OpenAI’s documentation, not an independent security assessment.

What to check before enabling it

The October 28 release note specifies iMessage on iPhone or RCS on Android; SMS is not supported. Users can turn texting off from the dot’s profile, and OpenAI says replying STOP also stops texts. Anyone considering the beta should review the connected dot’s memory, plugins and permissions, and account for Linq’s role in message delivery.