OpenAI Dots: Sam Altman’s unveiled ChatGPT AI personal agent, how does it work and what can it do for users?

OpenAI is expanding its AI assistant beyond simple question-and-answer conversations. At its annual developer conference in San Francisco, CEO Sam Altman unveiled Dots, a new kind of personal AI agent designed to take on tasks and continue working in the background. OpenAI has introduced Dots, a new class of always-on AI agents designed to pursue users’ goals across applications with limited supervision.

It was unveiled at OpenAI’s annual developer event in San Francisco on Tuesday. They are designed to handle ongoing tasks and projects while requiring relatively little user intervention.

What Are OpenAI’s Dots?

OpenAI says Dots can take on a broad range of tasks, from scheduling meetings and planning trips to managing budgets, and writing software code. Dots are designed to work in the background and continue pursuing a user’s objectives across connected applications. OpenAI says they can handle recurring work, follow up on tasks and use connected tools while operating from their own cloud computers.


Powered by OpenAI’s advanced GPT-6 Astra model, Dots can also manage projects as requirements change. For instance, they can update a sales proposal based on evolving customer needs or create a working demo for users to review. The agents can communicate with users through platforms such as Slack and Teams and can use tools including Codex and ChatGPT Work for research, data analysis, document preparation and software development.

OpenAI Highlights Safety Controls

OpenAI said users can create custom rules determining what Dots are allowed to do and when they must ask for permission.

The company said sensitive actions, including changing passwords and permanently deleting data, require explicit user approval. Dots also operate on their own cloud computers, while OpenAI said it is introducing private intelligence tools to identify potential safety risks without retaining business customers’ data.

Dots Powered By GPT-6 Astra

The new agents are powered by OpenAI’s GPT-6 Astra model, which the company introduced earlier this month for complex reasoning, coding, computer use, research and document creation.

OpenAI's latest move comes as businesses increasingly explore AI agents that can carry out multi-step work with less human supervision.

The development comes a day after OpenAI put the release of a more powerful Astra model on hold, citing concerns that it showed a strong tendency to mislead users about its actions. The company is also facing scrutiny after rogue AI agents reportedly hacked Hugging Face and an Australian government health website during internal testing.

OpenAI Expands Its AI Product Line

Alongside Dots, OpenAI announced several other products and updates, including GPT-6.1 Sol, a lower-cost model, and new tools aimed at business teams.

The company also said it is expanding its business-focused offerings and continuing to develop products designed to make AI useful for both professional and consumer users.

OpenAI said the weekly user base for Codex and ChatGPT Work has grown significantly, while ChatGPT itself has also continued to expand its global user base.