OpenAI launches Dots, always-on AI agents for complex tasks
OpenAI introduced Dots, proactive AI assistants designed to help users tackle complex projects and everyday tasks, on October 26, 2026, according to the company's official newsroom. The product represents a shift toward agents that can work independently while remaining under user supervision.
What Dots Do
Dots function as "always-on agents in ChatGPT" capable of continuing to make progress between conversations, according to OpenAI's Help Center. The agents can use their own computer access and integrate with connected apps, enabling them to research information, create content, and follow through on tasks without requiring users to restart conversations or provide constant direction.
OpenAI describes the assistants as "proactive assistants that can keep working across complex projects and everyday tasks" while maintaining user control over their operations OpenAI. This design philosophy prioritizes transparency and human oversight, distinguishing Dots from fully autonomous systems.
Market Positioning
The launch positions OpenAI alongside other AI vendors developing agentic capabilities. Dots ship as a public feature within ChatGPT, making them immediately accessible to the platform's existing user base. The product targets users managing multiple ongoing projects who would benefit from assistants that retain context and take initiative between sessions.
The timing reflects growing enterprise and consumer demand for AI systems that can operate with minimal intervention while preserving human decision-making authority. By embedding Dots directly into ChatGPT rather than releasing a standalone product, OpenAI leverages its existing distribution and user trust.
Technical Foundation
While official sources do not disclose detailed technical architecture, the product builds on OpenAI's existing capabilities in language understanding and task automation. Dots' ability to access external apps and maintain state across conversations requires integration with ChatGPT's broader ecosystem of plugins and data connections.
The announcement does not include specific pricing, availability by region, or feature tiers. OpenAI's Help Center confirms the product is operational and accessible through the ChatGPT interface, but the company has not released performance benchmarks, user adoption metrics, or comparative data against competing agent products.
What Users Control
A core promise of Dots is that users remain in control of agent actions. OpenAI emphasizes this principle in official descriptions, though specific guardrails—such as approval workflows, action limits, or audit trails—are not detailed in available public materials. This framing addresses longstanding concerns about autonomous system accountability and aligns with regulatory expectations around human oversight.
