OpenAI has launched Operator, an AI agent capable of autonomously performing tasks online, marking a significant step in advancing AI technology.
Available to US users on a $200 Pro subscription plan, Operator can handle tasks like travel bookings, shopping, and restaurant reservations. It operates via a dedicated browser, navigating websites, clicking buttons, and filling forms, mimicking human interaction.
Also Read: OpenAI Launches Operator: A New Autonomous AI Agent for Online Tasks
Operator uses OpenAI’s Computer-Using Agent (CUA) model, which combines GPT-4’s vision and reasoning abilities. It partners with platforms like DoorDash, eBay, and Uber, ensuring compliance with their terms. The agent requires user confirmation for actions like purchases and safeguards sensitive information by demanding manual input for critical details.
Although capable, Operator struggles with complex or highly customized tasks and requires user supervision for sensitive processes like banking. OpenAI has implemented limits on task numbers and introduced security measures to prevent misuse.
This development positions OpenAI in competition with companies like Google and Anthropic in the race to revolutionize digital interaction.
