US introduces next-generation AI agents: OpenAI's ChatGPT learns to work for the user

The American company OpenAI announced the launch of a new tool within the ChatGPT platform — a multifunctional AI agent that can perform tasks on a computer on behalf of a user. The new agent is available to subscribers of the paid Pro, Plus and Team plans and represents a technological step towards full automation of digital routines and business processes.

Developed on the basis of existing OpenAI solutions, the ChatGPT agent can not only analyze information and generate text responses, but also perform real actions: from managing calendars and creating presentations to writing and running program code. Thanks to integration with the virtual desktop, the agent is able to consistently move from analyzing tasks to performing them in practice.

Built-in functions include navigating websites, collecting and synthesizing information from multiple sources, and then creating short analytical reports. The user can interact with the agent through a familiar text query — just formulate the task in natural language.

Inside the system, the tool is implemented through the activation of agent mode, which allows ChatGPT to work in an interactive environment, simulating user actions. This approach significantly reduces the time required to complete typical tasks and increases productivity in both individual and corporate environments.

The launch of the AI agent marks a further development of the concept of intelligent automation, which is already being actively implemented in a number of industries-from marketing to engineering. Given the rapid spread of such solutions, it is likely that other developers will have similar features in the coming months, and the concept of "virtual assistant" will move to a new phase — agents that can perform complex tasks offline.

OpenAI continues its course to transform ChatGPT into not only a text generation platform, but also a full-fledged digital assistant of a new generation.