Remember When Google Said Your Data Stayed on Your Device?
Remember when tech companies used to put privacy promises right there in their settings pages, plain as day, where anyone could read them? It felt almost refreshing — a company spelling out exactly what it would and wouldn’t do with your data. Chrome did exactly that, with a clear statement that its on-device AI features ran locally, meaning your information never left your computer to reach Google’s servers. That promise is now gone. Not updated. Not clarified. Just quietly removed.
And that’s where our story gets interesting.
What Chrome Actually Said — and What It Says Now
Up until Chrome version 148.0, the browser included a specific line in its settings that read something like: “Chrome can use AI models that run directly on your device without sending your data to Google servers.” That sentence was a privacy assurance. It told users that certain AI features were self-contained — your device did the thinking, and Google stayed out of it.
That line has been removed. No fanfare, no announcement, no replacement statement explaining what changed or why. It’s simply gone.
For non-technical users, this might sound like a minor housekeeping edit. It isn’t. When a company removes a privacy claim without explanation, the reasonable question to ask is: did the underlying behavior change too, or just the promise?
The 4GB Download You Never Agreed To
The removed text isn’t even the most eyebrow-raising part of this story. Researcher and privacy advocate Alexander Hanff has reported that Google Chrome is silently installing a 4GB AI model onto users’ devices — without asking for consent first.
Think about what that means in practical terms. You open your browser one day, and somewhere on your hard drive, a 4-gigabyte file has appeared. You didn’t request it. You weren’t asked. Chrome just decided your device was a good place to store it.
Hanff has also suggested this practice may violate EU law, which sets strict rules around what software can be installed on a person’s device without their knowledge or explicit agreement. That’s a significant legal flag, not just a philosophical complaint about corporate overreach.
The Default-On Problem
There’s another layer here that deserves attention. After Chrome auto-updates — which happen automatically for most users — new features, including these AI additions, are switched on by default. You don’t opt in. You have to actively find the setting and opt out, assuming you even know the feature exists.
This is a pattern worth recognizing. Default-on settings are a deliberate design choice. Most people never change defaults. Tech companies know this. So when something is turned on automatically, the company is effectively making a decision on your behalf and counting on inertia to keep it that way.
For people who aren’t deep in the tech world — which is most people — this is genuinely hard to navigate. You’d need to know that a change happened, know where to look in Chrome’s settings, and know what you’re actually toggling off. That’s a lot of steps between you and a choice you should have been given upfront.
What This Means for Everyday Chrome Users
If you use Chrome, here’s what’s worth doing right now:
- Open Chrome’s settings and search for “AI” to see which features are currently enabled on your browser.
- Review any features labeled as on-device or experimental and decide whether you want them active.
- Check your device’s storage if you’re on a computer — a 4GB file appearing without your knowledge is something you have every right to investigate and remove.
- Keep an eye on updates from privacy researchers and EU regulators, since the legal questions around this practice are still developing.
Trust Is a Feature Too
There’s a broader point sitting underneath all of this. AI is being built into everything right now — browsers, operating systems, productivity tools, phones. That’s not inherently bad. AI features can be genuinely useful. But the way they’re being introduced matters enormously.
Removing a privacy promise without explanation, installing large files without consent, and defaulting new features to “on” without asking — these aren’t technical decisions. They’re trust decisions. And right now, Chrome is making some that don’t exactly inspire confidence.
Users deserve to know what’s running on their devices, what data those features touch, and where that data goes. A browser that once promised transparency on exactly this point has quietly walked that promise back. That’s worth paying attention to — especially as AI becomes less of a novelty and more of a permanent fixture in the tools we use every day.
🕒 Published: