OpenAI officially launched the long-awaited "ChatGPT Agent" platform on July 17, 2025, transforming its famous chatbot into a full-fledged AI agent capable of autonomously performing complex, multi-step tasks. This launch, actively discussed in the tech press on July 20, is the practical implementation of the previously announced "super assistant" concept and a direct response to similar developments from competitors like Google. As reported by leading publications Reuters and TechCrunch, ChatGPT Agent can take on comprehensive assignments that require interaction with the internet and other applications. For example, a user can give a command like: "Compare flight prices to London for next week, choose the best option, book it using my details, and add the trip to my Google Calendar." A key technological feature is that the agent operates in an isolated and secure cloud environment – a "virtual computer" – where it has access to its own web browser, code interpreter, and other tools. This allows it to safely browse websites, analyze information, fill out forms, and write and execute code to solve the given task. OpenAI specifically emphasizes that before performing critical or irreversible actions (such as making a purchase or sending an official email), the agent will always request explicit confirmation from the user. This release marks a fundamental shift from conversational AIs to action-oriented AI agents, changing the very paradigm of human-computer interaction: instead of performing the steps themselves, users delegate the entire task.
OpenAI Launches "ChatGPT Agent" Platform to Perform Complex Tasks
