The release coincides with rival AI company Anthropic unveiling its Citations feature.
OpenAI yesterday (23 January) released a research preview of its Operator artificial intelligence (AI) agent, launching in the US to its Pro subscribers.
AI rival Anthrophic also yesterday revealed its Citations feature, which aims to provide detailed references for its AI bot Claude’s responses.
OpenAI’s latest agent is powered by a new model called Computer-Using Agent (CUA), a combination of GPT-4o’s vision capabilities with “advanced reasoning through reinforcement learning”. According to the company, Operator can go to the web to perform tasks for its users.
“Using its own browser, it can look at a webpage and interact with it by typing, clicking and scrolling,” OpenAI explained. “Operator is one of our first agents, which are AIs capable of doing work for you independently – you give it a task and it will execute it.”
It claimed that Operator can handle a variety of “repetitive browser tasks”, such as filling out forms, ordering groceries and even creating memes.
“The ability to use the same interfaces and tools that humans interact with on a daily basis broadens the utility of AI, helping people save time on everyday tasks while opening up new engagement opportunities for businesses.”
The company further claimed that the research preview will allow it to learn from its users and the broader ecosystem and will therefore allow it to improve Operator.
Deleted data and limitations
OpenAI also revealed that it may store deleted Operator data for up to 90 days, even after a user manually deletes them. It added that its policies around data retention for Operator are designed to combat abuse.
And with any product in the early stages of its life cycle, OpenAI has acknowledged that Operator currently has some limitations.
“While it’s already capable of handling a wide range of tasks, it’s still learning, evolving and may make mistakes. For instance, it currently encounters challenges with complex interfaces like creating slideshows or managing calendars.”
As such, early user feedback will help to enhance the “accuracy, reliability and safety for Operator”, OpenAI asserted.
Earlier this week, recently appointed US president Donald Trump announced that private sector funding of $500bn would be invested into OpenAI’s infrastructure over the next four years.
The Stargate Project’s initial equity funders, which includes OpenAI, also includes Oracle, MGX and SoftBank, with Microsoft, Nvidia and Arm among the key technology partners.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
