OpenAI has unveiled a new feature called ChatGPT Agent — a virtual assistant capable of managing computers and executing a variety of tasks. The Agent is already available to ChatGPT Pro, Plus, and Team users, with a rollout for corporate accounts expected by the end of summer.

Unlike standard chatbots, the Agent can interact with calendars, create presentations, analyze documents, make online purchases, schedule meetings, and perform everyday actions such as booking restaurant tables. It has access to a virtual environment that includes a browser, terminal, file system, and other tools, allowing it to autonomously handle requests.

This feature is built on a new model that can work with multiple tools simultaneously. The development team has combined the efforts of the Deep Research and Operator projects, but unlike the latter, the Agent does not just consult or assist in decision-making — it executes actions independently in a virtual environment. Operator, on the other hand, involves human participation in task execution.

ChatGPT Agent can:

• Analyze calendars and organize events;
• Create documents, spreadsheets, and presentations;
• Work with user files;
• Conduct searches and make bookings via third-party services;
• Generate reports and process web data.

Before performing irreversible actions, such as sending emails or placing orders, the tool requires confirmation. Some functionalities, including transactions, are temporarily limited, and certain websites open in Watch Mode.

Currently, the feature is unavailable in the European Economic Area and Switzerland. OpenAI has not announced specific launch timelines for these regions.

The new tool emerges amid increasing competition in the AI agent sector. Google, Meta, Amazon, Anthropic, and startups like Klarna are already implementing similar solutions. According to OpenAI, ChatGPT Agent represents a step toward an AI model that not only consults but also performs real tasks.