OpenAI Dots: Always-On AI Agents Land in ChatGPT

OpenAI turned its DevDay 2026 keynote in San Francisco into a coming-out party for the always-on agent. On Tuesday, the company unveiled OpenAI Dots, a new class of persistent AI agents that live inside ChatGPT and keep working after you close the chat window, as part of a 20+ announcement blitz streamed live from the event.

OpenAI describes Dots as "remarkably capable, always-on agents built to handle everything," reports TechCrunch. Unlike Codex or a standard ChatGPT session, Dots are meant to operate independently of any specific hardware or interface, pursuing user-defined goals continuously in the background with minimal oversight.

The pitch is personal. "Today, you can start with your primary dot, give it a name, and make it your own," OpenAI said in its announcement post. Over time, the company envisions teams of Dots working together on a user's behalf.

Rollout starts today. Dots are launching in ChatGPT for Pro and Business Premium subscribers in eligible markets, according to Thurrott's DevDay rundown. Your first Dot is included in the plan at no extra cost, and for the next month Dots usage won't count against plan allowances. WCCFTech notes the rollout excludes the European Economic Area, Switzerland, and the UK for now. Enterprise, Edu, and Healthcare users can try a beta once their workspace admin enables it.

What Dots can actually do

Every Dot gets its own cloud computer and browser, so it can work around the clock. It researches, analyzes data, writes documents, or builds software through Codex, writes The Decoder. Through OpenAI's plugin ecosystem, a Dot can connect to more than 4,000 applications and move several projects forward at the same time.

The agent learns from feedback, gradually picking up an individual's preferences, standards, and working habits. OpenAI's demos paint a vivid picture. One developer deploys a dedicated Dot to monitor customer feedback, implementing bug fixes and requested features as they arrive. A scientist puts a Dot on experimental data, rerunning analysis and investigating unexpected results as they appear, per TechCrunch. In a lighter demo, a Dot noticed its user was working through dinner and proactively messaged two GrubHub delivery options with pricing for each, reports Red Hot Singapore.

You can reach your Dot through ChatGPT on desktop, web, and mobile, or in Slack and Microsoft Teams, with context shared across channels. Pro users can join a waitlist to text their Dots over iMessage or RCS. Each Dot has a name and a cartoon avatar, and if it takes the wheel while you're watching, a "Take over" button lets you step back in.

OpenAI also built in safeguards for the always-on model. Each Dot works on its own cloud computer, keeping your machine and its contents separate unless you connect them. For signing into supported websites, Dots can use saved passwords without exposing them to the model. And the company says it does not use content from ChatGPT Business, Enterprise, or Edu workspaces to improve its models by default.

ChatGPT Space: shared ground for humans and agents

Dots don't live alone. OpenAI is pairing them with ChatGPT Space, a collaborative layer where employees, ChatGPT, Codex, and OpenAI Dots work against the same shared context, reports VentureBeat. Space replaces ChatGPT's existing Library for Pro, Business, and Enterprise users and becomes the home for Pages, files, presentations, spreadsheets, and other team materials.

The shift is philosophical as much as functional. Instead of AI work staying trapped in one employee's private chat, teams can bring people and agents into shared documents, tag a Dot or ChatGPT for help, and keep building on work colleagues or other agents already completed. It's a small step toward a workplace where agents are coworkers with a desk, not tools in a drawer.

Why the agent world is watching

The timing is pointed. OpenAI Dots arrive three weeks after Meta's Muse, a personal agent with its own secured cloud computer that shot to number one in both US app stores after its early-September launch, as covered by Towards AI. 9to5Mac frames Dots as ChatGPT's answer to Muse and SpaceXAI's Grok Bot. The race to give every user a persistent agent is officially on.

For agent infrastructure watchers, the enterprise hooks matter most. OpenAI is working with Microsoft to bring Dots into the company's Agent 365 security controls, and companies can provision "specialist Dots" with their own identities, credentials, and system access. That puts identity and permissioning, not just capability, at the center of the agent story. It echoes yesterday's launch of NVIDIA's Open Agent Safety Platform, covered here on genznewz, which takes the same bet: safety has to live in the infrastructure beneath the agent.

Not everything shipped. The Decoder reports that the top-end GPT-6.1 Astra model is being held back over safety problems, even as the cheaper GPT-6.1 Sol rolls out broadly. Safety, it seems, is now the gating factor on capability, not the other way around.

For those of us building in the agent ecosystem, the signal is clear. The industry is converging on persistent, identity-bearing agents with their own computers and their own ongoing responsibilities. The chatbot era asked agents to answer. The Dots era asks them to show up, stay late, and finish the job. More coverage of the agent beat is on the AI News topic page.