Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

OpenAI launches Operator, an AI agent that performs tasks autonomously | TechCrunch

Jan 23, 2025 - techcrunch.com
OpenAI has launched a research preview of Operator, a general-purpose AI agent designed to automate tasks by taking control of a web browser. Initially available to U.S. users on ChatGPT’s $200 Pro subscription plan, Operator will eventually expand to more users and countries. The AI agent can perform tasks such as booking travel, making restaurant reservations, and shopping online by interacting with websites like a human, without needing developer-facing APIs. Operator is powered by a computer-using agent (CUA) that combines the vision capabilities of OpenAI’s GPT-4o model with advanced reasoning abilities. OpenAI is collaborating with companies like DoorDash and Uber to ensure compliance with terms of service agreements.

Operator requires user confirmation before finalizing tasks with external effects, such as submitting orders or sending emails, to ensure accuracy. However, OpenAI acknowledges that the CUA is not yet perfect and requires supervision for sensitive tasks like banking transactions. Operator represents OpenAI’s most ambitious attempt at creating an AI agent, surpassing the capabilities of previous virtual assistants like Siri or Alexa. This development marks a significant step in AI technology, aiming to transform how users interact with the internet and their devices by enabling AI to take actions rather than just processing information.

Key takeaways:

  • OpenAI has launched a research preview of Operator, an AI agent that can perform tasks independently using a web browser.
  • Operator is initially available to U.S. users on ChatGPT’s $200 Pro subscription plan, with plans to expand to other tiers and countries.
  • Operator can automate tasks like booking travel, making reservations, and shopping online, using a dedicated browser for task completion.
  • OpenAI emphasizes user confirmation for tasks with external effects and requires supervision for sensitive tasks, acknowledging that the CUA model is not yet perfect.
View Full Article

Comments (0)

Be the first to comment!