ChatGPT Dot Texting: iMessage, RCS, Setup and US Availability

OpenAI has started rolling out texting for ChatGPT dots, letting eligible US Pro users send their assistant messages through iMessage or RCS. The October 28 announcement adds text, photos and voice notes as ways to assign work, with proactive updates coming back through the same channel.

The useful change is continuity. A request does not have to begin and end inside the ChatGPT interface. But the preview has clear limits, especially for people outside the US or expecting an assistant that can text other people.

Who can use dot texting?

The official announcement describes a preview rolling out to Pro users in the US. OpenAI's help documentation adds that access is limited, so being a US Pro subscriber does not guarantee the feature is already enabled. Business and Enterprise workspaces are excluded from this beta.

That means this is not a global launch, and the announcement does not establish an India release date. Readers outside the supported rollout should not treat screenshots from another account as evidence that their own account has access.

OpenAI has not announced a separate texting price in these materials. Its help page says message and data rates may apply, which is different from promising that every part of the experience is free.

How to set up texting

OpenAI's follow-up post directs users to open their dot's profile and select Texting. The setup documentation says to connect the phone through the desktop app and follow the prompts, where the feature is available.

The company says the dot retains the same context and connected plugins across channels. In practical terms, this is another way to reach the existing assistant rather than a separate chatbot that needs the whole task explained again.

For example, someone could ask for an update on an ongoing research task while away from their laptop. The value comes from carrying the work into that shorter interaction, rather than making the user reopen the original workspace and reconstruct the conversation.

It cannot text your contacts

The texting channel is between the owner and their dot. OpenAI explicitly says only the owner can text that dot, and the dot can only text the owner. This release should not be read as permission or functionality for sending texts to clients, friends or coworkers.

OpenAI also says enabling texting does not give the dot or OpenAI access to the user's other text conversations. That is a useful distinction: connecting one assistant conversation is different from handing over an entire messaging inbox.

The help documentation still advises caution with sensitive information because texting uses a third-party provider. It also says replying STOP ends outgoing texts from the dot. Stopping messages and deleting the assistant's saved information are separate controls.

A more useful channel also needs restraint

For a proactive assistant, delivery is part of the product. A completed task has limited value if its result sits unnoticed in an app the user rarely opens. Texting can shorten that gap, particularly for a quick decision or an update that matters away from a desk.

The tradeoff is attention. A phone message feels more immediate than a background task update, so an assistant needs to distinguish useful interruptions from ordinary activity. Moving every status message into a text thread would make the channel less useful very quickly.

I think the important test is whether the assistant can bring back the one update that changes what someone should do next, while leaving the rest of the work in the background.

I'm Paras. I work at the intersection of AI, engineering and distribution.

I help AI teams turn product updates into clear technical stories that developers and founders want to read and share.

Book a 30-minute call.