In the previous post I talked about biometrics and in particular of facial recognition. Data extracted through these systems are surely more reliable than data got through other systems of surveillance so, even if still potentially dangerous, an automation of surveillance systems based on biometrics would be much more acceptable than a system which gets its data through tracking paths and interceptions.

A clear explanation on why these systems could be wrong and therefore extremely wrong is given by the computer security researcher Jacob Appelbaum.

“If anybody sees you on the subway, link your debit card to your metro card, right? And how do we feel? This is the concept […] called linkability. Take one piece of data and link it to another piece of data. So for example if you have your metro card and you have your debit card, you have those things and you can draw a line between them. So that’s like, not a scary thing, except your bank card is tied to everything else you do during the day.
So now they know, where you’re going, where you make purchases. So when they decide to target you they can recreate your exact steps, with your metro card and the debit card alone. Like literally where you go, where you buy. And potentially linking that data with other people on similar travel plans they can figurate who you talked to and who you met with.
When you then take cell phone data, which logs the location and you link up, purchasing data, metro card data and debit card. You start to get what you can call metadata in aggregate over the person’s life. And metadata in aggregate is content. It tells a story about you. Which is made up facts, but it’s not necessarily true. So for example, just before you where on the corner and all these data point to it, it doesn’t mean you committed the crime. “

Jacob Appelbaum, Occupy Wall Street training, NYC – from Citizenfour, film by Laura Poitras