OpenAI Introduced Dots Persistent AI Agents

The new automation tools are designed to operate independently across thousands of applications.

Updated on Sept. 29, 2026 in Artificial Intelligence

Isometric editorial illustration of modular server racks and data conduits, representing persistent autonomous AI infrastructure.
OpenAI has launched Dots, a new class of persistent AI agents designed to handle autonomous, long-term task management across cloud applications. AI Illustration. Upload story photo >

Live Poll

Do you trust persistent AI agents to handle your sensitive information securely?

OpenAI has officially launched Dots, a new class of persistent AI agents built on the GPT-6 Astra framework. These agents run on cloud environments to perform continuous research and task automation for users.

Why it matters

Dots represent a shift toward autonomous software that remains active even after users log off, enabling ongoing project management. OpenAI developed these agents to handle complex, multi-step workflows that require proactive research across various platforms.

Dots are powered by GPT-6 Astra and maintain persistent operations via cloud-based virtual browsers. Each agent can interact with over 4,000 distinct applications through plugin integration, with one agent included in current eligible plans.

The players

OpenAI

This research and deployment organization is a leading developer of artificial intelligence technologies.

The details

Users can set up and monitor their Dots through an Activity View interface available on ChatGPT desktop and browser platforms. While agents handle proactive research independently, any action that impacts a user account requires an automated review step to ensure security.

Timeline

  1. OpenAI introduced Dots on September 29, 2026, at DevDay.

The Tech Race

The introduction of persistent agents marks a transition from reactive chatbot interfaces to autonomous software capable of executing long-term tasks. This development places OpenAI in direct competition with other firms racing to dominate the emerging market for enterprise-grade AI agents.

ChatGPT Pro and Business Premium subscribers can now utilize these agents to automate repetitive research tasks without manual input. The integration with 4,000 apps streamlines workflows, though users should prepare to review and approve sensitive account-level actions.

The takeaway

Users should treat persistent agents as assistants that require oversight for sensitive operations, despite their ability to function independently. Implementing these tools allows for continuous task execution, shifting the user role from active operator to supervisor of automated workflows.

What happens next

OpenAI plans to launch enterprise-focused pilots for specialist Dots and will eventually introduce support for text messaging and increased agent capacity.

Further reading

For more background on current developments in the field, explore the Artificial Intelligence section.

Live Poll

Do you trust persistent AI agents to handle your sensitive information securely?