4 min learnCould 7, 2026 07:14 PM IST
Google’s Chrome browser is dealing with scrutiny after a safety researcher claimed that the browser quietly downloads and shops a big on-device AI mannequin with out clearly asking customers for permission.
The problem centres round Gemini Nano, Google’s light-weight synthetic intelligence mannequin designed to run straight on a tool as a substitute of relying fully on cloud servers. In response to researcher Alexander Hanff, additionally identified on-line as “That Privateness Man”, Chrome could robotically set up almost 4GB of AI-related recordsdata on appropriate Home windows and macOS techniques.
The invention has sparked considerations about privateness, storage utilization, and transparency, particularly since many customers didn’t know the recordsdata have been being saved on their units.
Hanff claims the AI mannequin is situated inside a folder named “OptGuideOnDeviceModel,” with a big file known as “weights.bin” believed to include the Gemini Nano mannequin weights. In response to the researcher, the recordsdata are robotically downloaded on techniques that meet Chrome’s {hardware} necessities, even when customers aren’t actively utilizing Gemini options.
The researcher additional mentioned that manually deleting the recordsdata doesn’t completely take away them. In some instances, Chrome reportedly downloads the AI mannequin once more after the browser restarts. A number of customers throughout on-line boards and tech communities have additionally mentioned discovering the recordsdata on their techniques over the previous yr.
Google has responded to the considerations by confirming that Gemini Nano has been a part of Chrome since 2024. The corporate says the mannequin powers a number of AI and security-related options, together with rip-off detection and developer APIs, whereas lowering the necessity to ship sure information to cloud-based servers.
In response to Google, the AI mannequin could robotically uninstall itself if a tool is operating low on assets. The corporate additionally said that it has began rolling out choices permitting customers to disable and take away the function straight from Chrome settings.
Story continues beneath this advert
Gemini Nano is Google’s smaller AI mannequin constructed particularly for on-device processing. Not like bigger cloud-based AI techniques, Nano runs regionally on a consumer’s pc, enabling quicker responses and doubtlessly enhancing privateness for sure duties.
Google has been quickly increasing Gemini AI throughout its ecosystem, together with Android smartphones, Search, Workspace apps, and Chrome. Within the browser, the AI mannequin is reportedly used for duties resembling rip-off detection, AI-assisted searching options, and developer instruments.
Right here’s a strategy to test
Customers who wish to test whether or not the AI mannequin is put in can kind “chrome://on-device-internals” into Chrome’s handle bar. This web page reportedly exhibits whether or not Gemini Nano is current and the way a lot space for storing it occupies.
If accessible, customers can disable the function via Chrome’s settings menu beneath the system part. Sure skilled customers have additionally advisable manually disabling Chrome flags referring to AI, however this will influence the browser’s efficiency or its AI-based capabilities.
Story continues beneath this advert
The problem signifies the broader controversy regarding using AI in client expertise merchandise. With technological corporations integrating synthetic intelligence into working techniques, net browsers, and different client units, privateness advocates insist that any large-scale AI mannequin shouldn’t be robotically put in with no consumer’s knowledgeable consent.
Regardless of Google’s assertion that Gemini Nano enhances consumer safety and creates participating synthetic intelligence experiences, the difficulty factors to growing societal considerations in regards to the hidden integration of AI in fashionable functions.
© IE On-line Media Providers Pvt Ltd

