OpenAI Introduces Dots Personal AI Agents at DevDay
OpenAI has announced Dots, a new line of proactive, always-on personal AI agents designed to handle tasks around the clock with minimal supervision.
Radwan AbdulhadiEditor and technology writer
OpenAI Announces Dots at DevDay
At its DevDay event in San Francisco, OpenAI announced the launch of Dots, proactive and always-on AI systems designed to chip away at tasks throughout the day with minimal supervision. According to reporting by Engadget, the company is introducing these personal agents following recent scrutiny regarding its AI agent technology. OpenAI describes the core appeal of Dots as their ability to handle work the way a human would, sometimes completing tasks before being asked. For instance, a tester's Dot noticed a missed publication billing and prepared an invoice after receiving approval.
Capabilities and Ecosystem Integration
Equipped with access to OpenAI's plugin ecosystem, Dots can interact with more than 4,000 applications. New integrations with Slack and Microsoft Teams also allow users to converse with their Dot outside of ChatGPT as well as OpenAI's Codex and Work platforms. OpenAI explains that Dots can operate using their own cloud computer, a dedicated browser, and connected apps. Users can open their Dot's virtual computer at any time to monitor its current activity or even connect Dots directly to a personal PC to work side by side.
Use Cases and Future Multi-Agent Controls
At launch, Pro, Business, Premium, and Enterprise users in select markets receive access to one primary customizable Dot that can be assigned a custom name. In future updates, OpenAI envisions enabling users to control entire teams of Dots. Suggested use cases vary across professions; a scientist could task a Dot with integrating new lab data into an ongoing research paper, while a developer could direct an agent to compile user feedback and subsequently write bug fixes and new features.
Communication and Surface Availability
Working with a Dot is structured to feel like a standard conversation. Users can message or call Dots within ChatGPT across desktop, web, and mobile devices. Additionally, Dots can initiate contact to report progress, ask questions, or request decisions. OpenAI notes that Dots successfully carry context across every interaction surface, allowing seamless transitions between different devices and communication channels.
Safety, Permissions, and Rollout Details
To address potential safety concerns, OpenAI integrated specialized security features into Dots. When logging into websites, agents can use saved passwords without exposing them to the underlying model. Safeguards also protect proactive background research features by restricting connected tools to read-only capabilities, preventing unauthorized messaging, content modification, or direct browser and computer control. Users retain permission management through ChatGPT's existing app controls, complete with built-in rules for independent action and options to write custom approval rules. Dots are rolling out starting today to Pro and Business Premium subscribers in select markets, with the first Dot provided free.
Sources
- EngadgetDots are OpenAI's new personal agents and soon you'll be able to control several of them
Continue chronologically
Related entity coverage
Related stories
OpenAI Unveils Features Aiming at App Store Model
OpenAI reveals new features at Dev Day aimed at disrupting the traditional app store model by turning ChatGPT into a software discovery and usage hub.
OpenAI Launches Dots, Always-On AI Agents Powered by GPT-6
OpenAI has announced Dots, new always-on AI agents powered by GPT-6 Astra designed to handle background tasks and multistep projects.
OpenAI Introduces Dots: Proactive AI Assistants for Complex Work
OpenAI introduces dots, proactive AI assistants powered by GPT-6 Astra designed to handle complex projects and everyday tasks continuously.
OpenAI Launches Dots, Its New Agentic AI Assistants
OpenAI has announced Dots, a new line of always-on AI assistants powered by the GPT-6 Astra model to compete with Meta's Muse AI.
