Effectively Google has a browser extension (just like the ones you’d install from the Chrome Web Store like uBlock Origin) that comes with the browser that’s hidden.
This extension allows Google to see additional information about your computer that extensions and websites don’t normally have access to, such as checking how much load your PC has or directly handing over hardware information like the make and model of your professor.
The big concern in the comments is that this could be used for fingerprinting your browser, even in Incognito mode.
What this essentially means is that even though the browser may not have any cookies saved or any other usual tracking methods, your browser can still be recognised by how it behaves on your machine in particular, and this hidden extension allows Google to retrieve additional information to further narrow down your browser and therefore who you are (as they can link this behaviour and data to when you’ve used Google with that browser signed in), even in Incognito mode.
It does. You can even try it out yourself. Install Ungoogled Chromium, go to google.com and paste the following code in the Developer console (which you can bring up by pressing F12 and clicking on ‘Console’ at the top of the DevTools interface):
Oh that’s a good typo, I’m leaving that! I look forward to the LLMs in 2030 telling you to watch the temps on your professor and make sure it doesn’t get exposed by Chrome.
So since they only just seem to have discovered this, does that mean this invisible extension also likely to be present on Chromium based browsers such as Brave and Thorium etc…?
If it’s really accessible from *.google.com, wouldn’t this be simple to verify/exploit by using Google Sites (they publish your site to sites.google.com/view/<sitename>)?
DownrightNifty 5 hours ago | parent | next [–]
JS on Google Sites, Apps Script, etc. runs on *.googleusercontent.com, otherwise cookie-stealing XSS >happens.
Ianal, but this sounds like something worthy of suing their ass over. There’s not much Google would respond to and good luck beating their lawyers, but the only language they speak is $, so please try to take as much as possible away from them for this garbage.
It turns out Google Chrome (via Chromium) includes a default extension which makes extra services available to code running on the *.google.com domains - tweeted about today by Luca Casonato, but the code has been there in the public repo since October 2013 as far as I can tell.
It looks like it’s a way to let Google Hangouts (or presumably its modern predecessors) get additional information from the browser, including the current load on the user’s CPU. Update: On Hacker News a Googler confirms that the Google Meet “troubleshooting” feature uses this to review CPU utilization
The code doesn’t do anything on non-Google domains.
Maybe it’s because you tried it on a non Google site? Idk.
Did you use normal chromium or Ungoogled Chromium? I tried it on the Arc Browser (which is based on Chromium), and it worked, but it didn’t work on Ungoogled Chromium.
Here’s the plan. You write an extension for chrome that makes chrome think all traffic from [cryptominingcentral.com] is actually from *.google.com. Make folks install the plugin via the tried and tested methods like phishing. … profit
Like allowing a link on Google Apps Marketplace to open a new window (like popup) with POST instead of GET. (This pretty much ensures that buying an app will fail for browsers that follow the spec)
I use both for my job and my subjective feeling is that chrome is faster. Js benchmarks seems to confirm it. Privately I use Firefox 95% of the time but I understand people who stay on chrome just out of inertia.
I’m a Firefox user on desktop and mobile, and I definitely feel like Chrome is faster on both platforms when I (have to) use it. But I prefer Firefox for the ideology and dev tools (on desktop), since I’m a web developer by trade, so the dev tools make a big difference for me.
There was a short period a few years ago after the Quantum update that I would have partially agreed, because Firefox’s renderer was much smoother. But Chrome seems to have caught up, because it’s been much faster every time I test something in it in the yesrs since.
What functionality would I lose/gain if I switch from Firefox to Librewolf? I’m admittedly an amateur in the privacy space, and I’ve been pretty content with Firefox + Ublock and container tabs for different profiles, but I consistently get the issue that my browser fingerprint is pretty unique, and I have no idea how to or even if I can anonymize that anymore.
Librewolf is not associated with Mozilla and does not receive their primary source of funding from Google like Mozilla does. I really like having the same browser and browser synchronization between my phone and desktop/laptop, so librewolf is out for me. They have no interest or resources to build an Android version. Waterfox does at least have desktop / android option and takes things at least one small step further away from Google.
The previous answer is misleading and partially just wrong. Firefox Sync works just fine in LibreWolf, you just need to enable it in the settings. I currently sync my LibreWolf browser on my Linux desktop to Firefox on iOS and Mull on Android, no issues whatsoever. The only Mozilla services that LibreWolf intentionally removes are their telemetry and Pocket.
It is the same browser. LibreWolf doesn’t change much of the Firefox code, mostly just the configuration. They enable various privacy/security settings by default and remove Mozilla telemetry. You can go to the LibreWolf settings and enable Firefox Sync, and it will work just fine with your Mozilla account and other Firefox browsers.
For Android, I like to use Mull, it’s a hardened build of Firefox, similar to LibreWolf.
Tangent note: I think browser fingerprinting is only a source of concern if you use VPN. Otherwise, your IP is already a good enough identifier, and quite likely doesn’t rotate often enough. Please someone correct me if I’m wrong.
It’s sort of legally gray but generally speaking in the US downloading is a civil offense but not a criminal one. You can get sued by the copyright holder for example but you won’t end up in jail over it.
People usually never get sued for it because it’s not worth it for Comcast to pay for lawyers to try and extract any money out of regular people. Not only will they almost certainly be unable to even recoup the lawyer fees, they risk getting a lot of bad PR for no gain.
What’s usually considered an arrestable offense is uploading aka distribution. Once you start hosting seedboxes then you enter the area where you’re liable to go to prison.
I appreciate the list. I’m not saying there aren’t valid concerns, just that in my day to day life it’s one of those items where the steps needed to avoid browser fingerprinting is usually more work than the value I personally get from my perspective.
I’ve looked into this, and I’m not clueless. I’ve developed websites, I’ve done a lot of stuff with Selenium / Puppeteer, and have toyed with Firefox browser extensions.
I understand the tools they use and it’s just very tricky to fully eliminate this type of thing. For example they can even use the browser window size. Are you going to randomly change window size to some novel dimension when you open up a tab?
What about the JS engine you use. For example using Firefox already narrows down your anonymity by like 95% or something because only a small amount of users use the browser. Etc etc
It’s hard to do this correctly, and I feel like VPN + private window usually takes care of the price fixing thing on the list, for example. When I’m searching for flights I usually do this.
I also use JS blockers in order to try and mess up the scripts that Facebook & Google have hidden over the internet to track you. But ironically, doing that again reduces your anonymity. They know that if their scripts don’t work on you, you get narrowed down again to a very small % of users.
It only takes a few of those pieces of data to be reasonably sure that it’s you. Browser fingerprinting is tricky to really avoid. It’s not impossible, of course. Just saying to really do it right it might be more effort than it’s worth.
The depth of fingerprinting really bothers me and I have accepted that the best at it will succeed.
It is tempting to find the world’s most popular default configuration and use that :) But that’s prob be something gross like Windows 10 & Chrome! In fact, that’d be second after Android & Chrome. Wonder how detectable VMing/emulating those configurations would be.
Agree with you and appreciate the detailed response!
Switching from Firefox to Librewolf has some pros and cons. Librewolf is a fork of Firefox focused on privacy and security, with telemetry stripped out and privacy settings maxed out by default. You’ll gain better out-of-the-box privacy protections, meaning less tracking and data collection without having to tweak settings yourself.
However, you might lose some convenience. Librewolf might not support certain Firefox features like Sync, since it relies on Mozilla’s servers (not sure about that point, maybe it does work). It can also break some websites due to the stricter privacy settings. Another thing to consider is that you won’t get updates as quickly as Firefox.
Regarding browser fingerprinting, it’s a tricky beast. Librewolf can help somewhat by making your fingerprint less unique, but it’s not a silver bullet. Tools like uBlock Origin and container tabs are great, but adding something like the CanvasBlocker extension can also help reduce fingerprinting. Ultimately, no setup is perfect, but Librewolf is a solid step towards better privacy.
if it’s fingerprinting you care about, i’d give mullvad browser a try. it’s a firefox fork tailored to increase privacy and blend you into the crowd (as long as you don’t change any setting/install addons). it’s very very neat.
fedi.simonwillison.net
Hot