If you opened ChatGPT after OpenAI DevDay and suddenly started seeing searches for “OpenAI dots,” “ChatGPT dots,” or simply “dots,” the confusion is understandable. OpenAI uses the lowercase name dots for a new class of always-on agents announced on September 29, 2026. They are not just a new chat theme or a row of decorative loading circles. They are designed to keep working on ongoing responsibilities, use connected tools, run recurring tasks, and return to you when a decision needs human judgment.
The fastest way to make sense of the term is to separate two different things that now share the word “dots.” OpenAI’s September 2026 product documentation describes dots as persistent agents powered by GPT-6 Astra. Separately, ChatGPT’s image-generation experience can show animated dots while an image is being created. Those loading dots are a visual effect, not the new agent product. OpenAI documented that animation in its ChatGPT release notes.
OpenAI describes dots as “always-on agents” built from its existing model and agent capabilities and powered by GPT-6 Astra. According to the company’s GPT-6 Astra System Card, each dot has its own cloud computer and browser, can use connected tools, can handle recurring work, and can follow up through ChatGPT and supported messaging channels. OpenAI also says dots may delegate work to subagents.
That is a meaningful change from the way many people use ChatGPT today. A normal conversation usually starts when you send a message. A dot is intended to take responsibility for a continuing goal, keep making progress between conversations, and bring you back in when it reaches a decision point or needs approval.
OpenAI’s getting-started guide for dots gives examples such as reviewing a calendar, researching a topic, running reminders, and carrying out recurring checks. The same guide says a dot can maintain its own activity view with work grouped under In progress, Scheduled, and Completed.
The timing explains most of the sudden interest. OpenAI held DevDay 2026 in San Francisco on September 29, 2026, and dots were introduced as part of that release cycle. The official OpenAI DevDay 2026 page confirms the event date and describes the conference as a venue for new APIs, tools, demos, and product announcements.
Searches also rise because rollout is gradual. Some users can see a dots setup flow while others cannot, and availability depends on plan, region, platform, and workspace controls. That naturally produces questions such as “Where is Dots in ChatGPT?”, “Why don’t I have OpenAI Dots?”, and “Is Dots different from ChatGPT agents?”
The clearest difference is persistence. A regular ChatGPT conversation responds to the prompts you send in that thread. A dot is designed around an ongoing responsibility rather than a single exchange.
This does not mean a dot has unlimited authority. OpenAI explicitly warns that dots can make mistakes and says users should review important work before relying on it.
As of September 30, 2026, OpenAI says dots are rolling out gradually rather than appearing to every account at once. The getting-started documentation says rollout has begun for eligible ChatGPT Pro users in supported markets, with exclusions listed for the European Economic Area, Switzerland, and the United Kingdom. OpenAI also lists Business Premium availability across supported ChatGPT regions and an Enterprise beta that requires workspace administrator enablement.
For Enterprise, the controls are more explicit. OpenAI’s workspace management guide says Use dots (Beta) is off by default. Local-computer access and custom rules are also separately controlled, and enabling dots does not automatically give the agent access to every connected app or website.
Start with the simple checks before assuming something is wrong with your account.
Dots are not described as a universal feature for every ChatGPT plan at launch. Check the current OpenAI help article for plan eligibility because rollout terms can change quickly.
OpenAI’s launch documentation includes regional limits for some plans. If you are in an excluded or not-yet-supported region, the feature may not appear even if your plan is otherwise eligible.
OpenAI says dot creation starts on the ChatGPT desktop app or desktop web. Mobile access may let you talk to an existing dot when that access is available, but the launch documentation says you cannot currently create a dot on mobile and that dots are not supported on mobile web.
If you use ChatGPT through a company workspace, an administrator may need to enable dots. Enterprise admins can also separately control cloud browser use, cloud network access, cloud-computer use, local-computer access, and connected-app behavior.
OpenAI says access may take several days to reach an otherwise eligible account. That means two people with the same plan may not receive the feature at exactly the same time.
Connected apps can give a dot information needed to complete work, but access is still limited by the permissions of the connected account and the controls configured in ChatGPT. OpenAI says dots can use connected information proactively and can form memories from that information, even when you have not just asked a new question.
That makes permission review especially important. For workplace deployments, OpenAI separates app availability, provider authorization, permitted actions, and approval requirements. In other words, turning on dots is not the same thing as granting blanket access to every file, mailbox, calendar, or external service.
Sometimes, but not without boundaries. The launch documentation describes several rule behaviors: take action without asking, take action if pre-approved, ask before taking action, or hand the task off to the user. Sensitive activities can require direct user involvement regardless of a custom rule.
OpenAI’s dots privacy and safety guidance says especially sensitive actions, such as changing a password or transferring money, may require the user to take over. Other operations, such as permanently deleting data or installing software, may require approval. These controls matter because a persistent agent can touch more of a workflow than a one-off chatbot response.
Not by default. OpenAI says every dot has its own cloud computer. Local-computer access is optional and starts turned off. If enabled through the desktop app, it can allow the dot to work with local files or local tools, subject to the permissions you grant and, in managed workspaces, the policies set by administrators.
This distinction is worth checking before troubleshooting. If a dot can browse in its cloud environment but cannot reach a file stored only on your laptop, that does not necessarily mean the feature is broken. The local machine is a separate access path.
OpenAI says dots can follow up through multiple channels, but individual channel availability is narrower than the broad product description may suggest. At launch, the getting-started guide says texting is a limited beta for some Pro users in the United States. It also says a dot cannot initiate phone calls at launch. Slack connections are supported where available, while enterprise messaging behavior can depend on workspace permissions.
Because these capabilities are actively rolling out, the safest approach is to treat the current OpenAI help pages as the source of truth rather than relying on screenshots or social posts from another user’s account.
Use this quick self-check:
If all you see is a few animated circles while ChatGPT is generating an image, you are probably seeing the image-generation loading experience documented in the ChatGPT release notes—not the dots agent product.
OpenAI dots are best understood as persistent ChatGPT agents that can keep responsibility for ongoing work instead of waiting for a fresh prompt every time. The launch matters because it moves ChatGPT further from a request-and-response tool toward a system that can schedule work, use connected services, operate in a cloud computer, and return for decisions or approvals.
For users searching because the feature is missing, the most likely explanations are ordinary rollout constraints: plan, region, platform, or workspace permissions. For users who already have access, the first thing to review is not a flashy demo but the control layer—connected apps, approval rules, scheduled tasks, memory, and local-computer access. Those settings determine what a dot can actually do in your account.
Last checked: September 30, 2026. OpenAI says dots are rolling out gradually, so availability and launch terms may change. Check the linked OpenAI documentation for the most current status.
Plan an aurora australis outing with realistic forecast checks, location trade-offs, camera settings, smartphone options, and safety tips for southern skies.
Explore NASA Curiosity’s newly released Mars dawn panorama, its mysterious yardangs, unusual blue tones, and what scientists still need to learn.
Learn what a hurricane watch means, how it differs from a warning, and what to do next to protect your household, prepare to evacuate, or shelter safely.
Learn how to read NOAA aurora forecasts, choose a dark and clear viewing spot, time your outing, and decide when to wait or change plans.
Explore Jaguar’s newly unveiled Type 01 electric GT, its range estimates, charging claims, key features, planned arrival, and details buyers should verify.
Understand OpenAI’s October 2026 math research release, Lean proof checking, model-access limits, and practical implications for learners and researchers.
Jaguar Type 01 is a production-bound electric four-door GT, not just a concept. See what Jaguar confirms and what U.S. buyers still need to verify.
ASOS says names and contact details may have been accessed, but passwords and card data are not believed affected. Learn what customers should do after the unauthorized alert.
Find the best Amazon UK tech deals during Prime Big Deal Days, October 6–7, 2026. Compare product specs, price history, sellers, delivery, and returns.
Learn how to use a hurricane tracker for live storm updates, read forecast cones, watches, wind probabilities, arrival times, and local alerts safely.
