OpenAI’s dots keep working after the instruction ends
Each dot has a cloud computer and a browser. Enterprise access is a switch an administrator turns on.
SAN FRANCISCO — OpenAI launched dots at DevDay on 29 September. They are persistent agents powered by GPT-6 Astra. Each one has its own cloud computer and browser, can use connected apps, and is built to continue a task after the first instruction, including recurring work and hand-offs to other agents.
ChatGPT Pro and Business Premium users in eligible markets get a first dot without a separate subscription fee. Work still draws on the plan’s usage allowance. Enterprise, education and healthcare workspaces can join a beta if an administrator enables it. OpenAI also previewed specialist dots for roles such as legal and finance, with their own identities, and said it is working with Microsoft 365 on how those are administered.
TechCrunch reported that dots can reach an ecosystem of more than 4,000 apps. That is the product claim. The operating claim is narrower. A dot that holds a browser and a cloud machine can book, send and change things. The buyer’s document is the list of those actions, the spend ceiling, and the name of the person who owns an exception.
Until that list exists, a dot is a pilot, not a colleague. 54east’s page on operations agents is the matching brief. https://54east.ai/operations-agents/
