Google is reportedly developing an artificial intelligence agent, internally known as “Project Jarvis”, designed to perform various tasks in a web browser.
Subscribe to the Softonic newsletter and get the latest in tech, gaming, entertainment and deals right in your inbox.
Subscribe (it's FREE) ►According to The Information, this tool would be showcased in December along with the next version of Gemini, its large language model, and would allow users to perform actions directly in the browser without the need for APIs or screen recording techniques, simplifying online research and other activities. It would be enough to just give instructions, and the system would act automatically, performing tasks such as filling out forms, booking flights, or making purchases on the Internet.
The plan to showcase Jarvis in December would be subject to change, according to The Information. Additionally, the outlet claims that Google is considering launching it among a select group of users to test it before releasing it to the general public.
Google is not the only company developing AI agents. For example, Microsoft recently announced the launch of its own agents, capable of performing various tasks without supervision. On the other hand, Anthropic announced a few days ago an agent called “Computer Usage” capable of controlling the mouse cursor to perform basic tasks and automating all kinds of processes on a PC.