OpenAI has officially unveiled its latest innovation, Operator, a groundbreaking AI agent designed to perform tasks on the web independently.
Touted as a major leap forward for artificial intelligence, Operator allows users to delegate actions like booking travel, making reservations, and shopping online.
“Operator uses a model called Computer-Using Agent (CUA), built on GPT-4o to interpret screenshots and interact with sites using typical browser controls like a cursor and mouse. You describe the task (e.g., “Book a flight,” “Order groceries”), and Operator executes the necessary steps. If it encounters a challenge—like a CAPTCHA or password field—it will pause and prompt you to take over, ensuring you stay in control,” reads the company’s website.
Accessible through ChatGPT’s $200 Pro subscription plan, Operator is initially available to U.S. users, with plans to expand globally over time, reported TechCrunch.
Unlike traditional AI assistants, Operator leverages OpenAI’s advanced Computer-Using Agent (CUA) model, combining the vision abilities of GPT-4o with enhanced reasoning skills. This enables the agent to navigate websites as a human would, using buttons, menus, and forms without relying on developer-facing APIs.
Users can activate Operator through a dedicated browser interface, allowing the AI to perform tasks independently while providing transparency into its actions. Despite its capabilities, Operator remains cautious, requiring user confirmation for actions with possible significant consequences, such as submitting an order or sending emails.
OpenAI has partnered with companies like DoorDash, Instacart, and Priceline to ensure Operator adhere to service agreements and function smoothly on their platforms. The AI agent’s safeguards include a monitoring system that halts operations if suspicious activity is detected and regular updates to its protective measures.
However, Operator is not without limitations. Tasks involving sensitive information, such as credit card details or complex calendar management, still require active user participation.
Operator’s release comes amid growing interest in AI agents, which promise to shift how people interact with technology. Unlike tools that merely process information, AI agents like Operator are designed to perform actions autonomously, effectively bridging the gap between passive assistance and active execution.