OpenAI is testing a new always-on assistant called “O”and references to the unannounced feature briefly appeared online.
OpenAI hasn’t confirmed or denied that it’s working on a new always-on assistant, but some users have confirmed it spotted On
Source: Jake on X
This feature was listed alongside increased Work and Codex usage, maximum memory, 100GB of file storage, and early access to new features.
In addition, there is a dynamic configuration with reference to display_name: "o" next to one email_suffix: "-o"and this seems to suggest that the Assistant may have some form of email capability or identity at some point.
This makes sense for an always-on assistant that could continue to help you with your tasks when you’re not actively using ChatGPT, but we don’t know exactly what OpenAI has planned yet.
OpenAI could reveal more at DevDay
OpenAI has already confirmed its hosting DevDay 2026 on September 29th in San Francisco and a number of new features are expected to be announced.
The event would give us a closer look at what the teams are building, along with technical sessions, demos and new tools. So I wouldn’t be surprised if we heard more about “o” there.
There was also evidence of an internal “Aeon” system around custom agents for ChatGPT workspaces, but “o” appears to be a separate, consumer-focused experience.
