OpenAI CEO Sam Altman unveiled a new “always-on” artificial intelligence agent known as Dots during the company’s annual developer conference on Tuesday, marking a significant addition to the firm’s suite of AI tools. The agents are designed to proactively assist users by carrying out ongoing tasks, positioning them as direct competitors to Meta’s recently launched personal agent Muse, which gained rapid popularity following Meta’s own conference last week.

Altman described Dots as an AI helper that continuously supports users, likening the functionality to the imaginative AI assistants often depicted in movies. The agents aim to provide seamless, persistent assistance, enhancing user engagement through constant availability.

The announcement came one day after OpenAI decided to pause the rollout of a more advanced model due to concerns about security raised by internal researchers. These concerns contributed to broader industry discussions advocating a slowed pace of AI development to allow for improved safety measures. While Altman did not directly address these security issues during his keynote, he acknowledged in a question-and-answer session that the company is investing considerable resources into safety, security, and vigilance regarding the operation and monitoring of AI agents.

In addition to introducing the Dots agents, OpenAI also revealed an upgraded iteration of its GPT-6 Sol model, named GPT-6.1 Sol, and unveiled a new premium tier called “ultrafast,” which offers enhanced processing speeds. These releases reflect OpenAI’s ongoing efforts to enhance both the capability and responsiveness of its AI systems amid a rapidly evolving technological landscape.

The company’s latest developments highlight the growing competition in AI personal assistant technology and underscore the ongoing tension between advancing innovation and managing associated safety risks in the industry.