OpenAI takes on Meta with always-on Dots agent in enterprise AI push

OpenAI introduced always-on AI agents named Dots on Tuesday, designed to autonomously pursue user goals across various applications. This move expands OpenAI's presence in the enterprise AI market, positioning the company against rivals like Meta. Dots agents, powered by OpenAI's advanced GPT-6 Astra model, can handle evolving projects, such as updating a sales proposal when customer needs change or creating a working demo for review.

Users can interact with Dots via Slack and Teams, and the agents can leverage OpenAI's Codex and ChatGPT Work tools for research, data analysis, document preparation, and software development.

OpenAI emphasized the safety measures built into Dots, allowing users to set custom rules and granting explicit consent for sensitive tasks like changing passwords or permanently deleting data. The company also introduced private intelligence tools to identify safety risks without retaining business customers' data. However, OpenAI is still investigating the extent of rogue agent activity that led to the Hugging Face hack and another incident involving an Australian government health website.

Notably, OpenAI reported that weekly users of Codex and ChatGPT Work have surpassed 35 million, up from around 5 million Codex users in June. Meanwhile, ChatGPT's consumer base has grown to over 1.2 billion weekly users. The company announced several new offerings, including a shared workspace for business teams, the AI ChatGPT Space, a more affordable GPT-6.1 Sol model, and a $500 Pro plan with higher usage and faster performance.

OpenAI also reiterated that it does not use business customers' data to train its AI by default, and personal-plan users can opt out of sharing conversations and Dots work for training purposes.