ChatGPT Agent brings autonomous task handling to OpenAI users
Users can activate the Agent by selecting ‘agent mode’ in the ChatGPT interface, available to Pro, Plus, and Team subscribers.

OpenAI has launched the ChatGPT Agent, a feature that transforms ChatGPT from a conversational tool into a proactive digital assistant capable of performing complex, real-world tasks.
By activating ‘agent mode,’ users can instruct ChatGPT to handle activities such as booking restaurant reservations, ordering groceries, managing emails and creating presentations.
The Agent operates within a virtual browser environment, allowing it to interact with websites, fill out forms, and execute multi-step tasks autonomously.
However, this advancement builds upon OpenAI’s previous tool, Operator, which enabled AI-driven task execution. However, the ChatGPT Agent offers enhanced capabilities, including integration with third-party services like Gmail and Google Drive, allowing it to manage emails and documents seamlessly.
Users can monitor the Agent’s actions in real-time and intervene when necessary, particularly during tasks involving sensitive information.
While the ChatGPT Agent offers significant convenience, it also questions data privacy and security. OpenAI has implemented safety measures, such as requiring explicit user consent for sensitive actions and training the Agent to refuse risky or malicious requests.
Despite these precautions, concerns persist regarding handling personal information and access to third-party services. Users must review the Agent’s permissions and settings to ensure their data remains secure.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!