The Daily Tech TipSmarter Tech. A Brighter You. ⌨ Home
ai
ai

Can you now text your ChatGPT agent like a friend — and how do you turn on Dot texting safely?

By Adam Neely · Oct 29, 2026
Quick answer: On October 28, 2026, OpenAI began rolling out texting for its Dot agent as a beta preview for ChatGPT Pro users in the US only. It works over iMessage and RCS, not plain SMS — here is how setup works and why you should never text it banking details, passwords, or ID scans.

On October 28, 2026, OpenAI began rolling out texting for its Dot agent as a beta preview for ChatGPT Pro users in the US only. It works over iMessage and RCS, not plain SMS — here is how setup works and why you should never text it banking details, passwords, or ID scans.

Your phone buzzes and it is not a friend — it is your AI helper, texting you an update before you even thought to ask. Short answer: yes, for some people. On October 28, 2026, OpenAI began rolling out texting for Dot, ChatGPT's always-on personal agent introduced in late September.

It is a beta preview for ChatGPT Pro users in the United States only, with limited access, and it is not available in Business or Enterprise workspaces. It works over iMessage on iPhone and RCS on Android — plain SMS is not supported.Why it matters: texting puts the agent where you already live, in your messages. You can send your dot text messages, photos, and voice notes, hand it tasks, or ask it to watch something for you — and the dot can text you first with updates it thinks are relevant.

OpenAI says only the account owner can text their dot and the dot can only text its owner, setup does not give OpenAI access to your other text conversations, and memory and context carry between ChatGPT and the text thread.Set it up safely, step by step:1. Open your dot's profile in ChatGPT and choose Texting.2. Verify your phone number, then save the contact so you know it is really your dot.3.

Check your messaging app first: iPhone needs iMessage enabled, and Android needs RCS enabled.4. Start messaging — and when a task needs your approval, the dot sends a link that opens ChatGPT so you can review it there before anything happens.5. To stop outgoing messages, text STOP.Pro tip: treat these texts as less private than they feel.

OpenAI runs the messaging through a third-party provider called Linq, reports say the exchanges are not end-to-end encrypted, and OpenAI's own help guidance urges care with sensitive information — so never text banking details, passwords, or scans of an ID to a dot.Ad's take: Texting an agent like a friend is genuinely handy, and that is exactly why the guardrails matter. Convenience should never talk you into sharing the keys to your bank account.

Sources: gagadget
Open on the homepage

Ask AI about this tip

Send this tip to your favorite AI assistant for a plain-language walkthrough, follow-up questions, or help applying it to your device.

Was this helpful?
Home