What you need to know
Google installed an incredibly large artificial intelligence model on users’ devices without their knowledge, according to a shocking report.
Google’s 4GB setup has been leaked Twins Nano model on users’ computers That Privacy Man. Their blog post goes into fine detail; however, one of the main points of concern is that this file is downloaded without users’ knowledge. The article notes that this model exists to implement processes such as Help Me Write, AI-powered browsing features, and AI fraud detection while browsing.
Upon further inspection, Alexander Hanff (That Privacy Guy) says that this massive 4GB download is triggered when “Chrome’s AI features are enabled, and these features are enabled by default in recent versions of Chrome.”
Article continues below
For Chrome’s alleged “benefit”, this Gemini Nano model has been the subject of discussion among Windows users who have experienced a drastic reduction in their PC’s memory (4GB to be exact). Moreover, users cannot function fully without this unwanted installation. Hanff says that users who never interact with Chrome’s AI features will still download this AI model to their hard drives without their knowledge. Removing it, as Hanf says, requires an inordinate number of hoops to jump through.
Users will have to discover the secret way Google hides its AI model on your machine, but it will simply “re-download it silently in the next appropriate window…”
This is where things get dicey. Hanff reports that with Chrome 147 (which many should already have), there’s an AI mode button on the far right of your address bar. Users can open a new tab to see it appear. You may find that the AI model on this device is useful for this, to perform tasks that you may require it to do. Hanff says otherwise. The post says that “every query entered by the user is sent over the network to Google’s servers to be processed by Google’s embedded models.”
This development is causing a stir, given that Google never asked users (any users).if they wanted to download this 4GB AI model to their computer. On the one hand, Hanff likens this behavior to what Anthropic did with Claude Desktop. Furthermore, it may be illegal under UK and EEA law. This could be a violation of Article 5(3), Hanff added: “The 4 GB Gemini Nano weights file is data stored on the user’s terminal equipment. The user did not consent.”
We see these features, but did we need them?
Let’s get one thing out of the way: most of the functions that this AI model “helps with” are tools we have already been exposed to. Back in April, Google expanded its search assistant to more countries. This feature can pull content from web pages or videos you visit to help you plan trips or summarize long articles online. In addition, the presence of Nano Banana allows you to transform the images you see into something completely new.
This feature is even here if it’s a bit relevant. These are AI fraud detection settings have been available on Android and desktop for some time. To go along with my colleague Jay Bonggolto, we’re usually given “you’re in control” messages from Google when it comes to AI — most AI. But then you see a report like this and the question arises: us?
Android Central’s Take
I hate it. Right. I’ll give credit where credit is due, my colleague Nicholas Sutrich came up with this title and I couldn’t agree more. It’s 4GB of my storage space, my computer, it’s gone… Google and it’s an AI model? The answer is no, and the answer will always be no. I agree with Alexander Hanff, this is a breach of trust from a consumer perspective. They didn’t ask anyone if they had this giant AI model on their computer. He didn’t ask anyone if it was “good”. This has just been done and Google is keeping quiet about it.
