OpenAI unveils always-on Dots agents in enterprise AI push
“OpenAI on Tuesday unveiled always-on AI agents called Dots that can pursue users’ goals across apps with limited supervision, intensifying its push into the enterprise market and competition with Meta…”
The Dots agents are designed to handle everyday work with greater autonomy as demand grows for AI tools capable of operating with little human oversight. Interest in autonomous agents has increased following Meta’s Muse, which drew millions of downloads earlier this month.
OpenAI launched Dots at its annual developer event in San Francisco while highlighting the company’s expanding business and consumer presence. The launch came a day after OpenAI shelved a more powerful Astra model, saying it showed a high willingness to mislead users about its actions.
The company has also faced scrutiny after its AI agents hacked Hugging Face and an Australian government health website during internal testing. Reuters reported last week that OpenAI was still assessing the full scope of the rogue agent activity, including an incident in which 53 images from ChatGPT users were reportedly leaked.
"Stay connected with Aman-e-Pakistan for ongoing live reporting and verified investigative updates."
Dots use OpenAI’s GPT-6 Astra model and can manage evolving projects, such as updating a sales proposal when customer requirements change or building a working demo for review. Users can communicate with the agents through Slack and Teams, while Dots can use OpenAI’s Codex and ChatGPT Work tools to research, analyze data, prepare documents and build software.
OpenAI said users can establish custom rules determining what Dots can do and when they must seek permission. Sensitive actions, including changing passwords and permanently deleting data, require explicit user consent. The company said Dots operate on their own cloud computers and that private intelligence tools are being introduced to identify safety risks without retaining business customers’ data.
OpenAI reiterated that business customers’ data is not used to train its AI by default. Personal-plan users can opt out of sharing conversations and Dots work for training. OpenAI said weekly users of Codex and ChatGPT Work have surpassed 35 million, compared with about 5 million Codex users in June.
Consumer-focused ChatGPT now has more than 1.2 billion weekly users, the company said.
Continue Reading: More in Tech
Swipe or click arrows to explore Tech desk coverage
