Google inadvertently confirmed the existence of advanced AI, “Jarvis,” following a leak revealing the tech giant’s latest AI developments. The leaked documents, originally published on Google’s own cloud platform, were quickly removed, but not before revealing an AI agent capable of autonomously browsing the internet and retrieving information.

This leak comes a few days after OpenAIown o1 model leakwhich exposed a new model that can analyze images and access tools such as web search and data analysis. It also comes less than a week after release Search ChatGPT and is similar to Anthropic’s latest beta project that allows Claude takes control of a user’s computer through a series of update screenshots.

Conformable informationsources close to Google suggest that Jarvis was intended for an internal pilot test before a potential public release. However, the unplanned reveal has fueled speculation about Google’s timetable for the official AI unveiling.

For all we know, Jarvis could visit websites, make online purchases with your permission, and even fill out forms. The Chrome-based AI, it would be the biggest update since the browser launched in 2008.

Unlike existing AI tools that rely on preloaded data, Jarvis is designed for search the web giving users the ability to retrieve data in real time, bypassing traditional search engines. Its functionality appears to be an extension of Bard AI from Google, combining natural language understanding with advanced data mining capabilities. This move toward real-time data responses is a shift from traditional ones chatbotswho are usually limited by their training dates and can only answer questions based on information up to a certain date.