OpenAI’s Dot messaging preview lets eligible US Pro users text their agents.
Users can send their agents texts, photos and voice notes, while receiving proactive updates through iMessage or RCS.
OpenAI has begun rolling out a Dot texting preview for Pro users in the United States. The feature lets users send instructions to their personal agents through texts, photos and voice notes.
Agents can also send proactive updates about assigned tasks. The rollout follows a promise OpenAI made when it introduced Dots nearly a month earlier.
How to set up Dot texting
To activate the feature, users need to open their Dot’s profile in ChatGPT and select Texting. iPhone users must enable iMessage, while Android users need RCS messaging.
Once connected, users can message their agent with a task or ask it to monitor something. The agent can then send relevant updates, giving users another way to track work they have assigned.
OpenAI says information a Dot saves to its memory carries across ChatGPT and text conversations. This means the agent retains that saved context when users switch between the two channels.
Messaging support and security
OpenAI partnered with Linq to support the messaging feature. The company also says it conducted extensive penetration testing to check its security.
In a statement on X, OpenAI said only the user can text their own Dot. Similarly, the agent can send messages only to that user.
The feature remains a preview for US Pro users. Dots are not currently available to subscribers on OpenAI’s Plus tier.
Competition grows in personal agents
Meanwhile, Anthropic entered the personal agent market last week with Claude Duet. Access currently requires joining a waitlist, which is open only to people in the United States.
Anthropic has said it plans to offer Duet to subscribers on its $20-a-month Pro plan. OpenAI’s latest rollout, meanwhile, adds a messaging option for eligible Dot users without extending access to Plus subscribers.
