OpenAI’s operator can browse the internet and perform actions on your behalf

January 24, 202512:39 PM EST No comments

OpenAI, the ChatGPT creatorhas released a new feature named Operator. This feature allows users to perform tasks and access the internet on their own. For now, it’s only for Pro users in the U.S.

OpenAI Operator uses AI for webpages to interact with them through typing, clicking and scrolling. The OpenAI Operator is still in preview mode and has some limitations. However, it will change and evolve based on feedback from users. The operator can do work for the user on its own. Give it a task and it will complete it.

The Operator can take actions by interacting with screenshots using a mouse or keyboard. This allows it to perform actions without the need for any custom API integrations.

OpenAI also says that the Operator is able to self-correct when it encounters errors or challenges. It can give control to the user if it gets stuck or needs assistance.

The ‘Computer Using Agent (CUA),’ is powered by a Computer-Using Agent’ model. This uses GPT-4o vision capabilities and offers advanced reason through reinforcement learning. The CUA can also be trained to interact with graphical interfaces such as buttons, menus and text fields that people see on their screens. It can perform actions such as ordering groceries, completing forms and creating memes. Operator only works within its own browser.

OpenAI said it was starting small, launching this feature for Pro Users in the U.S. and will take what it learns before expanding further. It’s not clear if the company plans to expand beyond the U.S. Learn more about Operator here.

Shutterstock is the credit for the header image.

OpenAI is the source.