Microsoft is running one of the largest corporate espionage operations in modern history.

Every time any of LinkedIn’s one billion users visits linkedin.com, hidden code searches their computer for installed software, collects the results, and transmits them to LinkedIn’s servers and to third-party companies including an American-Israeli cybersecurity firm.

The user is never asked. Never told. LinkedIn’s privacy policy does not mention it.

Because LinkedIn knows each user’s real name, employer, and job title, it is not searching anonymous visitors. It is searching identified people at identified companies. Millions of companies. Every day. All over the world.

you are viewing a single comment's thread
view the rest of the comments
[–] 140 points 5 months ago* (last edited 5 months ago) (39 children)

This is straight up misinformation. First off, it's perfectly legal.

LinkedIn does browser fingerprinting. It's the same thing Google and Meta do. It's how Google Ads is shifting to a post-adblocker revenue stream.

Browser fingerprints show fonts used, audio codecs, WebGL render data, processor, operating system - enough that if you add up several factors together, it makes a statistically unique fingerprint. it does NOT scan applications on your computer. It can't. It DOES scan which browser extensions you have running (if they affect page loading).

If you check your email and then close that and go to Google in an incognito window and search for porn - Google will fucking know what you're looking at. Gmail and all Google apps all fingerprint, and then you'll notice how Google ads trackers are on most sites online? Yep. That's how they track you.

Use a VPN? Use an ad blocker? Great - Google doesn't care. Google can track your fingerprint.

See your own fingerprint - check how it know it's you visit after visit.

https://fingerprint.com/

https://coveryourtracks.eff.org/

https://amiunique.org/

  • source
  • hideshow 39 child comments
  • [–] 41 points 5 months ago (6 children)

    it does NOT scan applications on your computer

    technically browser extensions are considered applications under EU's GDPR

    It DOES scan which browser extensions you have running (if they affect page loading).

    as per their report:

    Why two detection methods

    Method Technique What it catches
    AED fetch() against known resource paths Extensions that are merely installed, even if they inject nothing into the current page
    Spectroscopy Full DOM tree walk Extensions that actively modify the page, even if they are not in LinkedIn’s hardcoded list
  • source
  • parent
  • hideshow 6 child comments
  • [–] 17 points 5 months ago (4 children)

    it's misleading to say its searching your computer tho...? this invokes the thought of LinkedIn getting to rifle through your files like it has access to ~/Documents/ or smth.

    but yeah tracking you over the internet is similarly bad

  • source
  • parent
  • hideshow 4 child comments
  • [–] 8 points 5 months ago (3 children)

    it's misleading to say its searching your computer tho...?

    Wait, your browser extensions aren't on your computer?

  • source
  • parent
  • hideshow 3 child comments
  • [–] 12 points 5 months ago

    It's misleading because saying "search the computer" implies a breadth of scan that isn't present. That's like saying a website "searches the computer" to grab cookies generated by that site; technically true but worded to be misleading.

    To be clear this is bad, but it's important to be clear when explaining why it is bad to avoid creating resentment when the person you are explaining it to looks deeper into it themself and finds that it's not as bad as your explanation was implying.

  • source
  • parent
  • [–] 1 point 5 months ago* (1 child)

    I believe the point they’re trying to make is that they have access to APIs which describe particular software on your PC. You can argue based on the fact that, yes, the software is persisted on your filesystem. However, the API they access brokers [meta]data about the software. It’s not a filesystem API. If I add arbitrary files to an extension directory under my browsers path for extension persistence, they probably cannot see those arbitrary files unless the extension is built to allow it.

    There is a big difference between having direct and broad read access to the filesystem, versus the much smaller volume of data they can infer about your filesystem using APIs for browser extension data.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 5 months ago

    While browser extensions are considered apps under the GDPR, the headline is intentionally misleading. LinkedIn isn't "Illegally Searching your Computer." It's asking the browser for all the info it's maximally able to give up. We do need to define browser extensions in a way that doesn't use fear as clickbait to make it sound like LinkedIn has greater access to a device than it really has.

    And thanks for the correction on AED, I had seen another analysis a couple weeks back and I didn't recall correctly what was being collected.

  • source
  • parent
  • [–] 11 points 5 months ago

    I think the argument is that since some of the extensions that are probed can be political in nature, which can reveal political identity, which is potentially unlawful in the EU. However, it really needs to be up to a judge to make a decision on that.

    In general what they're doing is legal, and the BrowserGate people are using niggling little details, a handful of extensions out of the 6000 probed, to justify this argument. I couldn't say, especially as someone from outside the EU, whether this is actually illegal or not, but it's definitely in a nebulous area at the moment.

    Though I agree it's sensationalized in terms of claiming it's "searching your computer" and doing "corporate espionage."

  • source
  • parent
  • [–] 10 points 5 months ago

    There is literally a section called Why it's illegal.

    LinkedIn’s scan reveals the religious beliefs, political opinions, disabilities, and job search activity of identified individuals. LinkedIn scans for extensions that identify practicing Muslims, extensions that reveal political orientation, extensions built for neurodivergent users, and 509 job search tools that expose who is secretly looking for work on the very platform where their current employer can see their profile.

    Under EU law, this category of data is not regulated. It is prohibited. LinkedIn has no consent, no disclosure, and no legal basis. Its privacy policy does not mention any of this.

  • source
  • parent
  • [–] 10 points 5 months ago (9 children)

    Fonts, codecs, hardware, OS, extensions are all parts of a computer that never ever need to be transmitted to a website for it to function. Any information about them should be sandboxed, and if the website wants to display differently based on them, it can send static data or code in and get nothing back out.

  • source
  • parent
  • hideshow 9 child comments
  • [–] 9 points 5 months ago (7 children)

    I'm pretty sure for fonts they can tell because they have different widths, which affects page layout, which can be measured.

    There's a lot of stuff like that.

    Best would be make it illegal and give the law teeth. Solving it technically will always be an arms race.

  • source
  • parent
  • hideshow 7 child comments
  • [–] 2 points 5 months ago (6 children)

    Yeah, they can very easily get all of that right now. But functionally there's no good reason for any browser to let them. Page layout should be a one-way operation that doesn't allow information back through.

  • source
  • parent
  • hideshow 6 child comments
  • [–] 4 points 5 months ago (5 children)

    You'd have to kill a lot of JavaScript and CSS for that to work, and then a lot of legitimate function goes away.

    Done much web development work?

  • source
  • parent
  • hideshow 5 child comments
  • [–] 2 points 5 months ago* (2 children)

    You don't have to kill much functionality at all. Scripts that need to access that data should simply live in a sandbox with no network access. They can still do full computational layout.

    I have done exclusively web development work.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 1 point 5 months ago (1 child)

    So you're going to make it illegal to call getBoundingClientRect and then pass that information to fetch through any mechanism?

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 5 months ago*

    Essentially yes. Basically, think of two JS sandboxes that can manipulate the same DOM. One can make requests, but cannot retrieve local layout data. The other can get layout data, but not make requests. Both can set layout data.

    Web developers can use the former 99% of the time, and the latter for more precise work.

  • source
  • parent
  • [–] 1 point 5 months ago* (1 child)

    Surely functionality affecting display can be standardized to the point of making them useless for fingerprints? I don’t really care what font my browser uses, as long as I don’t notice it. Similarly, other details should either be randomized, mocked, jittered, or outright blocked. Fingerprinting only works because they’re operating in a rather non-adversarial space. The weakness with their current approach is the huge set of variables, which I’m sure we can leverage to reduce the algorithms determinism.

    We can either all appear the same, or appear completely unique every time. Either approach should work.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 5 months ago

    I don't know a lot about how fingerprinting works, but some of what i've read is pretty insidious. Some things could probably be obfuscated, but some of what the trackers use has legitimate purposes as well. Your application may serve different content based on the screen size, or fall back to an older library if such-and-such API isn't supported.

    Personally I'd rather make targeting advertising and tracking illegal, and gut the whole thing to avoid the arms race.

  • source
  • parent
  • [–] 4 points 5 months ago (15 children)
  • [–] 1 point 5 months ago (13 children)

    I have NoScript for JS tracking, but what do you use for fingerprint randomisation?

  • source
  • parent
  • hideshow 13 child comments
  • [+] 4 points 5 months ago* (last edited 2 weeks ago) (7 children)
  • [–] 3 points 5 months ago (6 children)

    I'm using canvas blocker and I'm still getting "Your browser has a unique fingerprint"

    What am I missing?

  • source
  • parent
  • hideshow 6 child comments
  • [+] 1 point 5 months ago* (last edited 2 weeks ago) (3 children)
  • [–] 1 point 5 months ago (2 children)

    Your link just opens this whole thread for me.
    Is there something wrong on my end (using Summit app) or is your link broken?

  • source
  • parent
  • hideshow 2 child comments
  • [–] 1 point 5 months ago (1 child)

    I know that in Threadiverse 2 weeks is a very long time but I just checked out CanvasBlocker and for a moment I had a similar question

    On the page of CanvasBlocker addon there is a link to https://browserleaks.com/canvas.
    And once I installed the addon, the page started saying Uniqueness: 100% (The signature is unique to our database) which sounds bad. But if you refresh the page, you can see that the Signature changes everytime. And I think this is the core, everytime a page tries to fingerprint my canvas, it will get a different signature. Every time I open a page, from canvas vector, I will look like a different viewer

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 5 months ago (4 children)

    I made no effort to do that, im using the duckduckgo browser on my phone.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 2 points 5 months ago (3 children)

    Interesting, I also have the DDG browser but the test shows a unique fingerprint result. I don't think that I have tinkered with any settings and I haven't installed addons.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 5 points 5 months ago

    Some of the test sites don't differentiate between random and unique. They may see a randomized fingerprint as a plausible unique user, but it may be different the next time you visit. Other sites may detect that your browser has taken steps to randomize your fingerprint, and use that as an identifying piece of information on its own (power user vs average joe)

  • source
  • parent