Introduction to OpenAI Dots

OpenAI has officially announced Dots, a new category of always-on AI agents unveiled during the company's DevDay 2026 event in San Francisco. Unlike a traditional chatbot experience, these personal agents are designed to continuously crawl the web, connect to user apps, and complete multistep projects on a user's behalf.

Visually depicted as customizable blobs that users can personalize, Dots are powered by the startup's GPT-6 Astra model. A primary selling point for the software is its persistent design, allowing it to pull context from connected applications and adapt to your preferences over time.

Demonstrated Capabilities and Use Cases

In a demonstration shared by OpenAI, a user's Dot proactively reviewed a busy calendar, recognized a working dinner schedule, and messaged the user with two GrubHub food delivery options complete with pricing. After receiving the user's pick and ordering instructions, the agent managed the request. Another demo showed a user collaborating with a Dot to launch a new website.

Users can interact with their Dots through multiple interfaces, including ChatGPT, Slack, and Microsoft Teams. Furthermore, users on the Pro tier can join a waitlist to text with their agents directly via iMessage or RCS messaging on Android.

Rollout and Pricing Tiers

The rollout for Dots begins immediately for subscribers of ChatGPT's Pro tier, which costs $100 a month. OpenAI frequently launches cutting-edge features first for its highest subscription tier before expanding them to a broader audience over time. While initial support allows users to control a single Dot, the company expects to introduce multi-agent control capabilities eventually.

Enterprise Focus and Specialist Agents

The introduction of Dots coincides with a stronger push by OpenAI into the enterprise client market. During his keynote appearance, CEO Sam Altman announced specialized versions known as specialist Dots. These enterprise-focused agents are tuned for specific professional tasks such as accounting, email marketing, and legal analysis, reflecting a broader trend of integrating AI coworkers into business workflows.

Safety, Security, and Privacy Considerations

OpenAI has integrated safety guardrails into Dots, including a design that requires explicit approval before executing sensitive tasks like installing software or altering a password. Additionally, users can configure a Custom Rules tool to establish strict operational boundaries.

Despite these precautions, experts note that connecting personal data to autonomous agents carries inherent privacy and security risks. Users must weigh the utility of automated assistance against the possibility of unintended information sharing or manipulation by external actors. Furthermore, if an OpenAI account permits model training on user data, that setting will extend to interactions involving the agents.

Sources

• WIREDOpenAI’s Dots Are Always-On AI Agents—and Its Answer to Meta’s Muse