Dots brings proactive agents to users
OpenAI CEO Sam Altman introduced Dots, a new line of artificial intelligence agents designed to remain active and complete ongoing tasks proactively for users. He presented the system at the company’s annual developer conference in San Francisco, describing it as a highly capable assistant that would continuously support its user. The product will compete with Meta’s personal agent Muse, which became popular after Meta’s conference the previous week.
Unlike tools that wait for a new prompt, Dots is intended to work on a person’s behalf over time. Altman framed the agents as helpers inspired by the friendly artificial intelligence assistants shown in movies. The announcement formed part of a larger group of product updates, including GPT-6.1 Sol, an upgraded version of GPT-6 Sol, and a premium speed tier called Ultrafast.
Safety concerns shadow the new products
A day before the conference, OpenAI halted the rollout of a more advanced model after its researchers raised safety concerns. The decision followed wider calls within the technology industry to slow the development of artificial intelligence long enough for safeguards to catch up. Altman did not address the shelved model during his keynote, but he later said the company was increasing its investment in the safety, security and monitoring of AI agents.
During his remarks, Altman compared the AI boom to a renaissance rather than an industrial revolution. He argued that the technology should increase people’s ability to create, learn, discover and expand knowledge instead of turning them into parts of an automated system. He also acknowledged that some areas of life should not be automated.
Technology companies adopt voluntary oversight
The OpenAI announcements came shortly before leaders from six major AI companies signed a voluntary agreement with President Donald Trump to supervise their own development practices. The accord calls for strong internal controls, reviews by independent external auditors and board committees that would evaluate reports from both internal and outside reviewers.
Although Altman did not attend the White House signing, OpenAI president and co-founder Greg Brockman participated alongside leaders from Anthropic, Amazon, Nvidia, Tesla, SpaceX and Microsoft. The agreement arrived as several figures in Silicon Valley continued asking the government to establish safeguards, while Trump resisted greater federal oversight of artificial intelligence.
References
Huamani, K. (2026, September 29). Altman unveils ‘always-on’ AI agent after OpenAI shelves model over safety concerns. AP News. https://apnews.com/article/sam-altman-openai-conference-dots-agent-77b6b8888145869206996d7509d24256