OpenAI has announced a significant update to ChatGPT Operator — an agent that can perform browser actions on behalf of the user. According to OpenAI’s vice president of research Jerry Tworek, Operator, which is still in preview, will soon become “a very useful tool.
Operator is a GPT agent that can automate routine tasks on the Internet: filling out forms, placing orders, searching for information, etc. The service appeared in January 2025 as a feature for Pro users for $ 200. Although the tool currently shows low performance, OpenAI is already preparing significant improvements. In answers to Reddit AMA Tworek hinted that Operator will be able to perform actions on the user’s computer and combine several tools into a single platform. This may be part of a global update to GPT-5, the company’s next flagship model, which will combine all services in a single interface.
OpenAI is actively exploring the integration of agent technologies with language models. Previous attempts to create an autonomous agent (such as AutoGPT) have shown the potential for task automation. At the same time, competitors such as Google with Project Astra are also moving towards integrating actions in the real environment. Operator is OpenAI’s first commercial attempt to implement “AI-out” with practical benefits for everyday tasks.
The expected update to ChatGPT Operator could radically change the way a user interacts with the Internet and PC by delegating routine actions to an AI agent. This is OpenAI’s strategic step towards creating a single platform ecosystem where GPT models not only respond, but also act. The key question is how reliable, secure and context-aware this automation will become.
61 
61 
75 