OpenAI launches Dots as safety concerns delay advanced AI model

OpenAI unveiled its Dots AI agent and other upgrades at its developer conference. The launch came amid fresh scrutiny after the company paused another model over safety concerns.

In Short

• Developers were shown agents built to handle ongoing tasks proactively
• OpenAI also introduced GPT-6.1 Sol and a premium Ultrafast speed tier
• Researchers' internal objections prompted the company to halt another model's release

OpenAI chief executive Sam Altman on Tuesday introduced a new artificial intelligence agent called Dots at the company’s annual developer conference, a day after OpenAI stopped the rollout of a more advanced model because of security concerns raised by its researchers.

At the event, OpenAI announced a series of products and updates, including the new agents designed to carry out ongoing tasks proactively for users. The announcements came as leading figures from the AI industry were meeting US President Donald Trump at the White House, amid a wider debate over safety rules and government oversight for the technology.

Altman described Dot agents as “like an AI helper that always has your back, inspired by the cool versions of what we all watched in movies growing up”. OpenAI’s agents are competing with Meta’s personal agent Muse, which surged in popularity after Meta’s conference last week.

On Monday, OpenAI said it was delaying the release of another model after concerns were raised internally by its researchers. The move came against the backdrop of wider calls within the industry to slow the pace of AI development so that safety measures can keep up.

Altman did not refer to the security concerns around the shelved model during his keynote address. However, during a question-and-answer session, he said the company was putting more investment into the safety, security and monitoring of AI agents.

He also said the AI boom should be seen more as a renaissance than an industrial revolution. “The best version of AI is not about making people cogs in a giant machine, ever whirring faster and faster. There are some parts of life that we cannot and should not automate,” Altman said. “AI should be about giving people more power over their own lives, more tools to create, learn, discover, expand knowledge.”

Among the other announcements, the company unveiled GPT-6.1 Sol, an upgraded version of GPT-6 Sol, and a premium speed tier called Ultrafast.

OpenAI’s announcements came as many of Altman’s industry peers gathered in Washington on Tuesday. While several leaders in Silicon Valley have urged the government to help put guardrails in place for AI, Trump has resisted calls for greater government oversight of the technology.

Altman was not at the White House meeting, but OpenAI president and co-founder Greg Brockman attended, along with Anthropic’s Dario Amodei, Amazon chief Jeff Bezos, Nvidia’s Jensen Huang, Tesla and SpaceX’s Elon Musk, and Microsoft chief executive Satya Nadella. Several other industry leaders, senior administration officials and House Speaker Mike Johnson were also present.

In sum, OpenAI used its developer conference to launch Dots and announce fresh upgrades, even as the company faces questions over AI safety after pausing another model’s release and as the wider industry remains divided over regulation and oversight.

With PTI Inputs

- Ends