OpenAI launches 'dots' agent one day after cancelling model

OpenAI debuted an AI agent called "dots" on Tuesday, less than 24 hours after cancelling the launch of a new AI model over safety concerns.

CEO Sam Altman unveiled the product at the company's annual developer showcase in San Francisco. He described dots as "more ambitious" than ChatGPT and "a whole new way to work with AI".

The launch closes a turbulent two-day stretch for the company. OpenAI had announced it would scrap the launch of an artificial intelligence model over safety concerns, and then followed that decision with the release of an entirely new suite of AI tools.

Why does the timing matter?

The sequence puts OpenAI's internal safety review process on public display. The company cancelled one product on safety grounds and shipped another within a single news cycle, a pace that will invite scrutiny of how the lab balances rapid releases against the safety concerns it cited.

The move also escalates the race among major AI companies to build agents — systems that go beyond chatbots to carry out tasks. OpenAI's dots arrives weeks after Meta introduced its own artificially intelligent agent, Muse. The two products now mark the front line of a competition to define what consumers and developers expect from autonomous AI tools.

What did Altman actually say?

At the developer showcase, Altman framed dots not as an increment to ChatGPT but as a step beyond it. He said the agent was:

• "more ambitious" than ChatGPT
• "a whole new way to work with AI"

That framing signals OpenAI's positioning of agents as the next product category after conversational chatbots, rather than a feature layered onto an existing model.

How does dots fit the Meta rivalry?

Meta introduced Muse, its own artificially intelligent agent, weeks before OpenAI's announcement. With dots, OpenAI has now answered on the same product category.

The stakes are commercial and strategic:

• Developers, the audience for Tuesday's showcase, are the constituency both companies are courting to build on their platforms
• Agents represent a shift from AI that answers questions to AI that acts, expanding what each company can sell
• The compressed timeline between Muse and dots shows how quickly the major labs are pushing agentic products to market

What about the cancelled model?

OpenAI's decision to scrap the launch of a new AI model over safety concerns preceded the dots announcement by less than a day. The company did not, in the information available so far, disclose which model was affected or detail the specific safety concerns behind the cancellation.

The juxtaposition is the story. A safety-driven cancellation followed immediately by a major product launch gives both critics and supporters of the company something to point to — evidence of a functioning safety process, or evidence of release pressure, depending on the reading.

What comes next?

The details of what dots can do, its availability, and how OpenAI addressed the safety concerns behind the cancelled model will determine whether Tuesday's launch reads as disciplined iteration or as pace winning out over caution. With Meta's Muse already in the field, the agent race has effectively begun.