OpenAI follows GPT-5.6 with ChatGPT Work, an agent that runs your apps
A day after GPT-5.6 went GA, OpenAI shipped ChatGPT Work: an in-ChatGPT agent that works across Slack, Teams, Drive, and email for hours at a stretch, metered like Codex.
One day after GPT-5.6 went generally available, OpenAI shipped ChatGPT Work: an agent inside ChatGPT, powered by GPT-5.6, that connects to Slack, Microsoft Teams, Google Drive, SharePoint, email, calendars, CRMs, and project trackers, and can stay on a task for hours. It rolls out to Pro, Enterprise, and Edu first, with Plus and Business over the following days, on web, mobile, and an updated desktop app for Mac and Windows.
The pricing model is the detail to notice. Work is metered like Codex rather than billed like a chat: complex tasks eat more of your plan’s included usage, and enterprise admins get spend controls and per-user limits. That is what an agent product looks like when the vendor expects it to run long enough to cost real money. In the same housecleaning, OpenAI killed its Atlas browser after eight months and folded the pieces back into ChatGPT, per The Decoder.
If your work lives in those connected apps, this is worth a real trial with a task you actually care about, not a demo prompt. Just go in knowing the meter is running, and that an agent with your email, your files, and hours of autonomy is a trust decision, not just a feature toggle.