OpenAI launches Dots, always-on AI agents for ChatGPT Pro users

In short: OpenAI has announced Dots, “always-on” AI agents that can keep working in the background and help with multistep tasks.

What happened

OpenAI introduced Dots at its DevDay 2026 event in San Francisco. The company describes Dots as AI agents, meaning they can take on tasks for you instead of only answering a single question at a time.

Unlike a regular chatbot, Dots are designed to keep checking for updates and moving a task forward after you assign it. Think of it like a helpful assistant who stays on the job, rather than someone who only responds when you speak and then stops.

In OpenAI’s demos, one Dot noticed a user was scheduled to work through dinner and suggested two GrubHub delivery options with prices. In another example, a user worked with a Dot on steps involved in launching a new website. OpenAI says Dots can also use context from connected apps and learn a user’s preferences over time.

Users can message Dots through ChatGPT, Slack, and Microsoft Teams. Pro users can also join a waitlist to text their Dot using iMessage or RCS on Android (a newer type of texting that can work more like chat apps).

The rollout starts today for ChatGPT Pro subscribers, which costs $100 per month. Users start with one Dot, and OpenAI is expected to add support for multiple agents later.

Dots are also designed to ask for explicit approval before sensitive actions, such as installing software or changing a password. Users can set boundaries using a Custom Rules tool.

Why it matters

More people are trying personal AI helpers that do things, not just talk. If these tools work as promised, they could save time on everyday chores, but they also raise practical questions about privacy, app access, and how much control users really have.

Source: Wired