OpenAI Launches Always-On AI Agent Dots
The new assistant can keep working between conversations, use connected apps and carry out recurring tasks through its own cloud computer.
Topics
News
- Apple Tightens Mac Access Rules as AI Agents Gain Reach
- Trump Puts Intelligence Chief Jay Clayton in Charge of AI Task Force
- IBM Widens AI, Quantum Research With IIT Bombay, IISc
- AI Dispatch | AI’s Big Bets Face a Reality Check
- Pentagon Plans Four-Star Command for AI, Drone Warfare
- OpenAI, Synopsys Build AI Model That Could Slash Chip Design Time
OpenAI has launched Dots, an always-on AI agent that can continue working after users leave a conversation, extending the company’s push beyond chatbots that respond to prompts.
Chief Executive Sam Altman unveiled Dots at OpenAI’s DevDay in San Francisco on Tuesday, September 29, describing them as “remarkably capable agents that can handle really anything you can think of.”
Powered by GPT-6 Astra, each Dot operates through its own cloud computer and browser and can connect to more than 4,000 applications, according to OpenAI.
Users can assign ongoing responsibilities rather than individual tasks. Dots can conduct research, review calendars, monitor connected applications, run scheduled checks and seek approval when decisions require human judgment.
“You just give your dot a responsibility … and your dots will just get to work and keep working,” Altman said at the event.
Dots can communicate through ChatGPT, Slack and Microsoft Teams. They can also proactively review information from connected applications, although such background research is restricted to read-only access and cannot send messages or modify content.
Users can set rules determining which actions an agent may perform independently, which require approval and which must remain under human control. Access to a user’s local computer is optional and disabled by default.
The launch follows scrutiny of OpenAI’s agent strategy after incidents involving unintended actions during internal testing. The company also recently delayed GPT-6.1 Astra after the model failed to meet its safety threshold on issues including authorization and adherence to instructions.
Altman said OpenAI needed to keep alignment, monitoring, safety and security ahead of increasingly capable models while continuing to release useful products.
Dots are rolling out gradually to Pro and Business Premium subscribers in eligible markets. Enterprise customers can access a beta version when enabled by workspace administrators.
The first Dot is included in eligible subscriptions at no additional charge, although more intensive tasks remain subject to usage limits.
