OpenAI unveils always-on AI agent Dots as Altman sidesteps security questions

OpenAI chief executive Sam Altman used the company’s annual developer conference on Tuesday to launch Dots, a set of “remarkably capable, always-on” AI agents that work proactively on a user’s behalf. The agents take on Meta’s personal agent Muse, which gained traction after Meta’s own conference the previous week.

The launch came a day after OpenAI said it was holding back a new AI model because of security concerns raised by its own researchers, a decision Altman did not mention on stage. He instead framed the AI boom as a renaissance rather than an industrial revolution, saying some parts of life should not be automated.

OpenAI also announced GPT-6.1 Sol, an upgrade to its GPT-6 Sol model, and a premium speed tier called Ultrafast. While Altman was at the conference, OpenAI president Greg Brockman joined other industry leaders at the White House for a meeting with President Donald Trump, whose administration has resisted greater government oversight of AI.