OpenAI launches Dots, always-on AI agents in ChatGPT with their own cloud computers
SiliconANGLE Duncan Riley ● Covered by 22 sources
OpenAI put always-on agents into ChatGPT, and each one gets its own cloud computer. They can keep working across apps and channels, even when no one is watching.
Based on reporting by SiliconANGLE, Duncan Riley — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI used its DevDay conference to push ChatGPT past the old one-prompt-at-a-time rhythm. Its new Dots agents are meant to keep moving on a task after the user steps away, keep track of preferences, and even surface work before anyone asks for it. Each agent runs on GPT-6 Astra and gets its own cloud computer that can be opened to watch what it is doing.
The pitch is less chatbot, more persistent assistant with memory and reach. OpenAI says the agents can connect to more than 4,000 apps through plugins, and with permission they can also work directly on a user’s laptop. They can be reached in ChatGPT on desktop, web, or mobile, and also in Slack and Teams, with context carrying over when a project moves from one place to another.
OpenAI is already using the setup internally. In one example, a bug reported in Slack sends agents into investigation mode. In another, new designs are turned into working apps while staff stay focused on customer feedback. An outside tester got a small taste of the same thing when an agent noticed an unpaid invoice, prepared it, and waited for approval before sending it.
There’s a safety story wrapped around the convenience. When an agent has no assignment, it can go looking for work through what OpenAI calls proactive research, but those app connections are read-only then. Anything that touches an account or shares information has to clear auto-review first, and a monitoring system can pause or stop an agent over safety concerns. OpenAI also says users should review consequential work because the agents can still make mistakes.
Dots starts rolling out today to ChatGPT Pro and Business Premium subscribers, though some accounts may wait several days. Pro users in the European Economic Area, Switzerland, and the U.K. are left out for now. Enterprise customers on Edu and Healthcare plans can try a beta once an administrator turns it on, and OpenAI says one agent is included with Pro or Business Premium at no extra cost for the first month of use.
My take — AI-written commentary, not fact-checked reporting
This is the real bet: not a smarter chatbox, but a software worker that sticks around long enough to be useful. That’s also where the trouble starts, because the minute an agent can act across Slack, Teams, and a laptop, “helpful” and “risky” live in the same chair. The industry keeps selling autonomy like a feature; users will keep treating it like a liability until the defaults get much stricter.
Read more about this at: SiliconANGLE