OpenAI debuts Dots as persistent agentic rivals to Meta’s Muse

Dots function through a text-based interface and voice calls integrated into ChatGPT, with the ability to bridge communication across tools like Microsoft Teams and Slack. Users currently manage a single agent, though OpenAI intends to expand this to allow multiple simultaneous deployments. When assigned a project—such as a developer iterating on software or a creator editing video—the agent provides status updates and proactively suggests follow-up actions based on historical usage patterns.

Collaboration happens within ChatGPT Space, where teams can interact with these agents in real time. To address security concerns following the recent, controversial launch of Meta’s Muse, OpenAI has implemented custom safety protocols. These include user-defined permission blocks and an auto-review system that forces the agent to verify actions against specific compliance rules before execution. Access begins today for ChatGPT Pro and Business Premium subscribers, with specialized enterprise roles currently in testing.