ChatGPT Dots and Space: persistent agents, shared context

OpenAI's new always-on agents take on ongoing responsibilities, and teams share context in one space.

At DevDay 2026 on 29 September, OpenAI introduced Dots, always-on agents that take on ongoing responsibilities, and ChatGPT Space, a shared workspace where teams and agents work with the same context. Together they point to AI that works alongside you over days, not single chats.

Key facts

Announced at OpenAI DevDay 2026 (29 September 2026).
Dots are always-on agents with ongoing responsibilities, built on GPT-6 Astra.
ChatGPT Space is a shared workspace where teams and agents work with common context.
Dots come with Pro, Business Premium and Enterprise; Space with Pro, Business and Enterprise.

What happened

Alongside new models and developer tools, OpenAI used DevDay 2026 to show two products that change how people work with ChatGPT over time.

Dots are always-on agents that take on ongoing responsibilities. They are built on GPT-6 Astra and each one has its own computer in the cloud, so it works inside the apps you have connected. You give a Dot a goal and decide what it is allowed to do on its own. Dots are available with ChatGPT Pro, Business Premium and Enterprise.

ChatGPT Space is a shared workspace. Teams and agents work in the same place from shared knowledge, so a person joining a project, or an agent picking up a task, starts from the same common context as everyone else. Space is available with Pro, Business and Enterprise on desktop and web; on mobile, for now, pages can be read and shared.

OpenAI also expanded ChatGPT’s plugin system, letting developers build sidebar experiences, interactive panels and custom file viewers that can react to events.

Why it matters

The big shift in AI this autumn is from single answers to ongoing work. OpenAI’s Dots, Anthropic’s Claude Duet and similar products all try to keep context alive over days and weeks.

When AI remembers and continues, context becomes the most valuable ingredient. A shared space only helps if what is inside is clear: the goal, the constraints, what has been decided and what is still open. Teams that keep a short, up to date brief for each project will get far more out of persistent agents than teams that rely on long, messy chat histories.

The same habit helps humans too: new colleagues get up to speed faster, and handovers before holidays become painless.

How to try it

You can start building that habit today with our prompt Chat handoff brief. Paste a long conversation, a set of meeting notes or your project notes into the brackets, and say who the brief is for, for example “a colleague who joins the project next week” or “a new AI chat”.

The prompt turns the material into a compact brief: goal, work done so far, decisions taken, open questions and next steps. Read it and correct anything that is wrong, then save it at the top of your project folder or paste it as the first message in a new chat.

A good test: give the brief to a fresh chat and ask “what should happen next?”. If the answer matches your plan, the brief has captured the context well.

Try it

Move a long AI chat to a fresh one without losing context

Frequently asked questions

What are ChatGPT Dots?

Dots are always-on agents announced at OpenAI DevDay 2026 that take on ongoing responsibilities. You give a Dot a goal and decide what it can do on its own.

What is ChatGPT Space?

It is a shared workspace where teams and agents work with common context, so everyone involved in a project works from shared knowledge.

How do I get the most from persistent agents?

Keep a short, clear brief for each project with the goal, constraints, decisions and open questions, and update it when things change.

Sources