Facebook Doesn’t Eavesdrop on Your Mic. The Truth is Much Creepier
The technology and techniques Facebook and other big tech companies use to profile you are much more sophisticated than simply wiretapping you.
Yet again, the excellent podcast Reply All is dovetailing with a topic that is near and dear to our heart here at groovyPost: Facebook privacy. In their most recent episode, Alex and PJ tackle a hot topic: Is Facebook Spying on You?
Specifically, is Facebook listening to you through your microphone and then serving you up ads based on your conversations?
The episode is excellent, and I highly recommend you listen to the whole thing. The hosts of the show think they come up with a definitive answer that matches Facebook’s official answer: No, absolutely not.
But even then, they have trouble convincing people like you and me—and perhaps even themselves—that Facebook isn’t truly listening in.
The thing that I heard that immediately made me want to share the episode with all of you is this:
One of the strongest arguments that Facebook and Google and Apple and Microsoft don’t eavesdrop on your audio conversations via your phone is that they don’t need to. Because their methods for tracking you otherwise are so dang sophisticated, that it’s better than if they were listening to your conversations.
I’ll explain a few of their techniques in a bit. But I’ll tell you that hearing it was a bit unnerving, but ultimately not really surprising. It’s all stuff I sort of assumed and accepted they were doing. The revelation is that with all this aggregate data about me and you, they can do better than wiretapping us. They can almost read our minds. Or at least the algorithms can.
Remember that news story about when Target guessed a man’s teenage daughter was pregnant before anyone else in the household knew?
That was five years ago. The technology has only improved since then. And the techniques and software we know about are only the tip of the iceberg. See, the industry’s ability to collect more and more data about you and analyze and parse that more and more accurately and intelligently is growing every day.
The eerie, spot-on super relevant nature of the ads is just the algorithms doing an excellent job.
It’s not magic. It’s not espionage. It’s science and engineering.
As an example, here’s a quick highlight from the episode about the most well-known tracking method Facebook uses.
Facebook Pixel Tracks Your Activity All Around the Web
We all assume Facebook is watching us closely when we are logged in. But what about when we are just browsing around on the web? Turns out they are watching there, too. How? One of the ways is with Facebook Pixel, a tiny little tracking code that web developers install on their website. This Facebook Pixel is practically ubiquitous on the web, and it can tell Facebook a lot about your browsing habits. It tells them what pages you looked at, how long you looked at them, when you looked at them, what kind of device you looked at them from and from where you looked at them (e.g., at 9:30 AM from the office on your mobile device or at 2:00 AM from home on your laptop).
Stat counters have worked this way for many, many years. This has nothing to do with cookies, and you can’t prevent it from going incognito or using private browsing. Using a VPN to anonymize your web activity may throw them off your scent, as long as you don’t log in to any accounts with your current IP. At any rate, as long as your device can access Facebook.com, there’s a good chance that Facebook’s data bots can find you.
Here’s an excerpt from the podcast where they talk to Antonio Garcia Martinez, a former Facebook developer who was the grandfather of Facebook Pixel.
ALEX: So he wanted to figure out a way to keep tracking people after they left Facebook. Like, to be able to see what they were doing all across the internet. And so he developed this thing that’s now called Facebook Pixel, and it’s installed on millions of websites. So when you go to one of these sites with Facebook Pixel on it, it watches what you do and reports that information back to Facebook. It can see how long you linger on a certain webpage, it can see if you purchase something, it can see if you put something in your cart on a website and decide not to buy it. It’s kind of like an internet surveillance camera.
PJ: Got- so that’s why like, that’s why, like, when you look at a pair of shoes or whatever- it follows you around Facebook.
ALEX: It follows you around the internet. Right. There’s this app that I use called Ghostery that shows you if Pixel is on a site that you’re visiting. And it’ll also show you all the other ad trackers that are on that site. Like, if you go to the New York Times website, there may be 30 or 40 of these trackers.
PJ: Like, as soon as there’s an ad, you basically have to picture 30 or 40 like helpful friendly sales associates like following you around the store–
PJ: –trying to guess how much money’s in your wallet, like, guessing your like weight and age, and like being like, “Oh, he looked at the hooded sweatshirt, oh my god, OK, OK, write that down, write that down, he likes hoodies.”
Here’s the thing I want to add about this: this is just about Facebook Pixel. Facebook Pixel is something that web developers voluntarily put on their website to help them with their own Facebook ad campaigns. There are probably dozens of other ways that Facebook can cross-reference web statistics in order to build a more detailed profile of you. It’s sort of like if Facebook had a tiny spy camera on everything in the world that was blue. As long as something blue was in the room, they could be watching.
Try it out. Go to any web page and view the page source. Ctrl+F for “facebook.com” and see if anything pops up, particularly images or any other HTML tag with “src” in it.
What Else, Facebook?
I’m not going to summarize the entire podcast episode here, but you should definitely give it a listen. Some other topics they cover:
- Facebook (and other tech companies) buy personal data about your offline activities from companies like Equifax and shadier brokers. These brokers also buy information or outright own loyalty card programs for places like grocery stores, restaurants, etc.
- Facebook gathers up data from your friends and contacts in order to build a profile of you. In some cases, they may even serve you ads based on your contacts’ profile.
- Facebook categorizes you based on what they know about you. You can view this by going to your Facebook ad settings and clicking the Your Information tab and then clicking Your categories.
I’ll say that my “Your categories” page is fairly innocuous and actually sort of inaccurate. They got me on the easy stuff, but I don’t know that I’m very liberal or a frequent traveler. All the rest of the info here you could glean just from my groovyPost articles.
What concerns me more, hypothetically, is the underlying data that Facebook—and every other company—has on me. The industry has just so many pieces of the puzzle and so many ways of putting those pieces together, it should be no surprise that they know more about us than we realize and perhaps know things about us that we don’t even know about ourselves.
Right now, it’s fairly benign. It’s all just data being fed into a vast algorithm. It’s not like Mark Zuckerberg personally knows anything about me or cares. In fact, the developers probably don’t know exactly how their inputs turn into their outputs.
But imagine this.
Imagine someone hacked some data center and got all this raw data from every Facebook user—all the stuff that Facebook Pixel knows about you. And what if Wikileaks or someone put it into a nice searchable database. You type in a name, and you sort by “websites visited” or “purchase history” and you find a bunch of stuff you have no business knowing. You could find out how Facebook categorizes someone and why. You could find out who someone knows and how often they contact them.
Of course, much of that data would have a questionable pedigree. But it raises the question. And it would be out there for your friends, your enemies, your family, your employers, your creditors, law enforcement, etc. to see. And suddenly, you’d be the one answering questions about the data.
Were you really searching the web for divorce attorneys?
Why did you buy that book on gambling addiction?
Are you a closet Baltimore Ravens fan?
I don’t mean to be alarmist. And in fact, I don’t have any immediate plans to change my online behavior. I don’t plan to delete my Facebook account anytime soon. Still, I predict that we will be talking about this a lot more in the coming years.
Or at least I hope we will be.
How do you feel about big data advertising techniques and privacy implications? Did you know all this already? Do you think it’s a threat? Share your thoughts in the comments.
It goes far deeper than advertising. The shadow Gov wants all info on us they can get.
It’s a disaster.
We need to maximize TOR and its ilk. We need to implement severe AUTOMATIC penalties for asking for any information beyond what’s needed for the current discussion or conversation.
Speaking only for myself, I advocate a policy making it illegal to ask for any information other than what is needed for the specific item at hand. Example: a weather reporting service could reasonable need to know where I am at that moment, or what area I”m asking about. There is no excuse for its asking what I had for breakfast or what color my shoes are.
I agree wholeheartedly.
I think it should be illegal to lose the data. Like HIPAA and other privacy laws. That’ll motivate companies to protect it better. Or perhaps obfuscate the data somehow so if it does leak, it’s anonymized.
Michael R Edwards
I read George Orwelks book 1984. Facebook has a lot of comparisons to the government in the book. One concern I have would be a corrupt leader might gain access to this information. He/she would know who has guns to confiscate. A crook could hack the account and could better target his/hers robberies. They could determine who has alarms, merchandise worth stealing, guns and when the house would be vacant.
I think it is time for Congress to restrict this data compilation.