OpenAI introduces AI agent Operator for booking travel.
OpenAI has launched Operator, an AI agent available to ChatGPT Pro users in the United States through a $200 monthly plan at operator chatgpt com. Operator is designed to autonomously perform tasks, such as interacting with websites by typing, clicking, and scrolling.
OpenAI CEO Sam Altman has described this as “an early research preview,” with the system still having limitations. It is expected to evolve based on user feedback, and OpenAI plans to release more agents after Operator. Eventually, it will be available to ChatGPT Plus, Team, and Enterprise users worldwide.
Powered by a new model called Computer-Using Agent (CUA), built on GPT-40, Operator can “see” through screenshots and perform actions within a browser using a mouse and keyboard, allowing it to take actions on websites without needing custom API integrations. Users can prompt Operator through partner websites or a search engine like Google. While it can self-correct errors, it’s still in its early stages and not yet as effective as humans at web navigation.
The goal of Operator is to turn AI from a passive tool into an active participant in the digital ecosystem, improving task efficiency and creating innovative customer experiences that drive conversions. Early partners include companies from the travel and mobility sectors, such as Booking.com, Hipcamp, Tripadvisor, Myshortlets, Uber, and Priceline, aiming to meet real-world needs while adhering to industry standards.
Alyssa Ravasio, CEO of Hipcamp, expressed enthusiasm about using OpenAI’s technology to enhance outdoor experiences, highlighting the importance of nature in an increasingly tech-driven world.
Users can prompt Operator for specific tasks, but can also take control of the browser at any time. Operator is designed to ask users to take over for tasks that involve logins, payments, or CAPTCHAs. Additionally, users can personalize workflows for specific tasks and run multiple tasks simultaneously.
OpenAI emphasizes user safety, incorporating three key safeguards: ensuring users remain in control, protecting data privacy, and defending against malicious websites.
Looking forward, OpenAI plans to release the CUA model via API for developers to create their own agents. Future updates will include handling longer workflows and expanding access to users on other plans.
