Introducing dots | OpenAI
Short summary
OpenAI is releasing Dots, autonomous AI agents powered by GPT-6 Astra that run 24/7 to handle tasks independently or alongside you across ChatGPT, Slack, and Teams. They learn your preferences over time, connect to 4,000+ apps, and include built-in safeguards so you control what they do.
- Dots can work autonomously on their own cloud computer, access connected apps, and handle multiple projects simultaneously, or work directly alongside you on your device with permission
- The system includes granular controls: you can set custom rules for when dots act independently vs. require approval, review their work in real time, and certain sensitive actions like password changes always require your involvement
- Specialist dots for enterprises come with their own identity and credentials, integrating with Microsoft Agent 365, designed to handle defined responsibilities like procurement and customer support with team oversight
Full summary
OpenAI is launching Dots, AI agents powered by GPT-6 Astra that run continuously on their own cloud computers and can connect to over 4,000 apps through plugins. Unlike ChatGPT, which requires prompting, Dots work proactively on your behalf—investigating bugs in Slack, turning designs into working apps, or preparing pull requests with test videos. They learn your preferences and work style over time, accessible through ChatGPT, Slack, Teams, or voice calls. The rollout starts with a primary dot per user on Pro, Business Premium, and Enterprise plans in eligible markets, with plans to expand availability.
The control mechanisms matter here. You can inspect a dot's work in real-time, set custom rules for when it acts independently versus asks approval, and restrict its access to read-only operations for background "proactive research." Dots use saved passwords without exposing them to the model, and sensitive actions like password changes stay with you. Auto-review checks whether actions align with your instructions and safety requirements before execution. Dots have their own isolated cloud computers, keeping your personal devices separate unless you explicitly connect them.
Beyond personal use, OpenAI is previewing specialist dots designed for organizations—these have company-provisioned identities, credentials, and access to internal systems. A specialist dot might handle recurring responsibilities like processing expense reports or managing database queries, improving through feedback while staying confined to its assigned role. The company emphasizes that dots still make mistakes, so you need to review consequential work, and they're not training directly on proactive research or your dot's internal notes by default.