You're Being Watched

Posted January 13, 2010 in Internet

Deep in the lush, disorienting Bwindi Impenetrable Forest in Uganda, a silverback mountain gorilla and his wives patiently munch berries. Nine human beings watch, immobile. Every morning a ranger uncovers broken branches and excreted berries to find the nomadic group's newest camp. The eight awed white tourists in his tow quietly snap as many photographs as their cameras will allow, careful not to disturb the creatures. Only 650 of the animals exist in the world.

In the forest of the Impenetrable Internet, technology companies analyze queries and clicks in their own journeys to piece together migratory patterns and learn more about the elusive human race. There are half a million times more Internet users than mountain gorillas in the world, so you'd think it would be hard to single out any one person from the masses. But Internet users forget: every step they take is through these companies' territories.

The companies set up security cameras. They can watch all the human beings they like.

They probably know you're reading this.

Pick Google, for instance. Every time you search using Google, you show Google a "cookie": a face for it to recognize you by. You can disable cookies, but that's like wearing a ski mask to the grocery store: your efforts are silly because the shelf-stockers can still identify you by your "IP address"--the fact that you came in a blue '98 Honda Civic with license plate WH00P5. Without an IP address, you can't get to the Google website at all, but you can use a "proxy server" to disguise where you came from. Just like taking a bus to the grocery store, using a proxy server is a pain in the neck, takes time, is unfit for children and the elderly, and still gives part of your identity away because while Google might not recognize you, it knows which bus you took.

Every company that's big enough will watch how individuals behave, for the customers' own good. Companies research whether customers find what they want or whether they spend ages looking for bread in the baking aisle instead of the bakery aisle. Google passive-aggressively suggests that you angle for "kournikova" when you search for "anna kornikova", because dozens of men (and women?) have overcome the spelling hurdle already, alerting Google to the confusion. Amazon suggests that you buy the latest Harry Potter DVD to complement your new toaster, because other exemplary customers bought the two items together.

Google is privileged to be a gateway to other websites. You use it to browse to BBC News; you use it to find pictures of Anna Kournikova; and you use it to find directions to the nearest liquor store. Every time you find what you want on Google and go to that page, you tell Google your next Internet destination.

Google is even more privileged to be a major Internet advertiser. When you view a page with a Google ad, Google sees what page you're reading. You can "opt out" of this practice, but the technological joke is that the only way to opt out is to send a cookie to Google each time you see an ad.

Because Google knows what you search for and which webpages you read, it can build a "profile" of you. Several companies do this; Google is unique in its relative honesty about privacy concerns. Browse to http://www.google.com/ads/preferences to see your profile and freak yourself out. A half-dozen other companies probably know that same information about you.

I doubt anybody can imagine how much money can spring from a database of 500 million Internet users and their behaviour.

And people wonder why Google lets you search for free.

Let's pick on someone else. Twitter users often use http://tinyurl.com/, http://bit.ly/, http://j.mp/ or other such websites to fit web addresses into Twitter's 140-character limit. In goes an address like http://maps.google.ca/maps?f=d&source=s_d&saddr=76+9+Avenue,+New+York,+NY+10011,+United+States+(Google)&daddr=1600+Amphitheatre+Parkway,+Mountain+View,+CA+94043,+USA+(Google)&geocode=FXurbQId08aW-yFatxpZiXCuXCktkJYTv1nCiTEixjLqxYw7rQ%3BFa8DOwIdLyS5-CEW5-Xqa0glpilRbQR7ArqPgDEel6cyESl4DQ&hl=en&mra=pe&mrcr=0&sll=39.081877,-98.197183&sspn=32.847436,54.580078&ie=UTF8&ll=39.571822,-98.173828&spn=32.636524,54.580078&t=h&z=4 and out comes the Twitter-ready address of http://j.mp/4LAFwg, which now points to the same thing. Great, but now the company behind j.mp sees where Internet users browse, since every John Twitteruser tells j.mp where he's going when he follows a j.mp link. This information on where Twitter users browse is worth money, and j.mp sells it.

I haven't even mentioned GMail, Google DNS, the iTunes Store, Facebook, or the dozens of other websites you frequent. They all know what you're doing. They may say their privacy policies are legally binding, but you and every other average person won't understand the policies even if you read them. Besides, laws vary from country to country and money-loaded companies have a long tradition of breaking laws first and buying solutions later: it's brave to hope that your Internet footprints are private.

But darn it, all these websites are useful. How can you use them and keep a sense of privacy?

You can't do a thing. Clearing cookies may confuse smaller companies, but big ones can recognize you regardless. Anonymous proxy servers are complicated and inconvenient. If you're religious, you can try praying, but anyone who's worked in a corporation can tell you how futile prayers are against large, lawyer-laden institutions.

My advice: when you use websites, write emails, post photographs, tweet, and instant-message your friends, assume a bunch of people you've never met are watching your every move and keeping records. Because they are, and you can never find out for sure what they're doing with their studies of your behaviour.

Maybe this is how the mountain gorillas feel.

Disclosure: while I have worked at Google, this post shares no corporate secrets and I have no ill feeling toward the company. It's just a handy example for this post containing widely-available, public information.