The Operator AI agent was released last month in preview
It is powered by an agentic model with GPT-4o’s vision capabilities
Only ChatGPT Pro users can access the AI agent
OpenAI’s Operator comes with its own dedicated web browser
Photo Credit: Reuters
OpenAI began rolling out its artificial intelligence (AI) agent Operator in several countries on Friday. The agentic tool was so far available in preview to ChatGPT Pro users residing in the US, after the company introduced it in January. Equipped with a dedicated browser, the Operator is a general-purpose AI agent that can autonomously perform tasks online based on prompts given by the user. While the AI agent is now available in several major regions, the AI firm said it will not be available in European countries.
OpenAI's Operator AI Agent Finally Rolls Out
In a post on X (formerly known as Twitter), the official handle of OpenAI announced the rollout of the AI agent. The company said that Operator will now be available to ChatGPT Pro users in “Australia, Brazil, Canada, India, Japan, Singapore, South Korea, the UK, and most places ChatGPT is available.” Notably, the tool is already available to those residing in the US.
However, the company also added that European countries will have to wait a bit longer before they get access to the AI agent. “Still working on making Operator available in the EU, Switzerland, Norway, Liechtenstein & Iceland,” the post added.
At the time of launch, OpenAI explained that Operator is powered by the Computer-Using Agent (CUA), which is a special AI model that gets computer vision capabilities from GPT-4o, and advanced reasoning prowesses from an unnamed OpenAI model (which could be the o3 model). The AI agent was post-trained using reinforcement learning.
The CUA has agentic capabilities and can interact with graphical user interfaces (GUIs) such as buttons, menus, and text fields displayed on the screen. It also has a dedicated browser that allows it to execute actions in a separate space, freeing the human user to carry out other tasks.
Operator accepts both text and images as input. While completing tasks, it analyses the raw pixel data of the screen and performs actions using a virtual keyboard and mouse, also provided in the sandbox environment. The AI agent is claimed to perform multi-step tasks, navigate through errors, and adapt to unexpected scenarios.
OpenAI Reportedly Working On an AI-Powered Web Browser, to Compete With Google Chrome10 July 2025
Search
