The algorithm behind the Facebook data scandal sounds like something out of a sci-fi movie. It digs through the smallest, most random posts – those “likes” people click on while scrolling – to uncover private info about someone’s sexual orientation, race, gender, even intelligence and childhood experiences.
Just a handful of “likes” can strongly guess which party someone supports, their gender and whether their partner is likely male or female, hint at whether their parents stayed together growing up, and predict if they might be vulnerable to drug abuse. And it does all this without reading messages, posts, photos or any other personal stuff on Facebook.
Some findings seem less like high-tech analysis and more like clever internet detective work; after all, “liking” a political page is pretty much the same as putting up a campaign poster.
But around five years ago some psychology researchers showed you could actually figure out way more complicated traits by spotting patterns that people wouldn’t notice just by looking at profiles. A few seemingly random “likes” could build surprisingly detailed personality profiles.
For example, liking “curly fries” and Sephora makeup was linked to intelligence; Hello Kitty likes hinted at political views; and clicking “Being confused after waking up from naps” was connected to sexuality.
These were some odd but consistent links found in a 2013 paper published in the Proceedings of the National Academy of Sciences. “Few users were associated with ‘likes’ explicitly revealing their attributes. For example, less than 5% of users labelled as gay were connected with explicitly gay groups, such as No H8 Campaign,” said the peer-reviewed study.
Michal Kosinski, David Stillwell and Thore Graepel – the researchers behind it – saw how creepy this could get and warned about privacy issues. Back then Facebook made all “likes” public by default.
“The predictability of individual attributes from digital records of behaviour may have considerable negative implications because it can easily be applied to large numbers of people without their individual consent and without them noticing,” they said.
“Commercial companies, governmental institutions, or even your Facebook friends could use software to infer attributes such as intelligence, sexual orientation or political views that an individual may not have intended to share.”
For some folks though, this sounded like a business idea. By early 2014 Cambridge Analytica’s CEO Alexander Nix had teamed up with Aleksandr Kogan-a fellow Cambridge researcher-for a private venture that resembled Kosinski’s research but aimed at commercial use.
Quick Guide How the Cambridge Analytica story unfolded Show
In December 2016 while digging into the US presidential election Carole Cadwalladr stumbled upon data firm Cambridge Analytica – its secretive style and shady background didn’t match its bland academic-sounding name.
Her early research revealed billionaire Robert Mercer’s involvement in Trump’s campaign: his strategic fight against mainstream media plus funding politics tied loosely to Brexit too.
She spotted clues suggesting Cambridge Analytica might have broken Data Protection laws using data methods during elections. That story pushed Britain’s Electoral Commission and Information Commissioner’s Office into investigations covering Cambridge Analytica’s data use including links to Brexit votes. These probes are ongoing alongside wider inquiries into political data usage by ICO.
As she followed these complex manipulations involving both money and info laws Cadwalladr faced attacks online plus professional pushback from key figures involved.
The Leave. EU team tweeted an edited video falsely showing her getting attacked violently while Russia’s embassy accused The Observer newspaper that she worked for of bad journalism outright calling her work “a textbook example.”
Still though her rising profile gave whistleblowers faith she’d get their stories right for everyone else clearly too.
Her network grew bigger including former workers who regretted what they did plus academics lawyers worried about democracy being affected by tactics used by Cambridge Analytica and partners.
Now special prosecutor Robert Mueller is investigating Cambridge Analytica’s role in Trump’s win. UK probes are still running strong.
Thanks for your feedback.
The academic created a Facebook app with a personality quiz , paid for test takers through ads on sites like Amazon Mechanical Turk.
The app saved quiz answers , grabbed data from each user ’ s Facebook account – crucially also pulling info on their friends too.
They matched quiz results with each user ’ s profile aiming to find patterns & build an algorithm predicting traits in others based just on FB data. Friends ’ profiles helped test this formula & more importantly made it valuable politically. Aleksandr Kogan
To take part you needed both a Facebook account & be registered voter in US so millions could be matched against electoral rolls easily. From just 1 ,000 initial users called seeders , researchers collected info on roughly 160 ,000 profiles – about 160 per person. Later hundreds of thousands would become test subjects unlocking huge amounts of US voters ’ info.
This was super tempting but also probably illegal since Kogan wasn’t allowed permission for commercial uses – his approval from Facebook limited him strictly to academic purposes only when gathering large amounts of profile info.
Though back then apps were allowed to collect friends’ info only within Facebook itself meant mainly encouraging interaction among users selling or using this friend-data outside FB-like what happened here-was banned totally especially when used politically by CA.
Looks like UK data laws were probably broken too because those forbid selling or using personal details without proper consent-and that includes cases where consent is given one way but data gets used another way entirely.
Those paid testers agreed to terms letting their own info be gathered plus default FB settings let apps collect friends’ details unless changed-but none signed off on having their information turned into big political marketing databases or tools.
Kogan insists everything he did was above board saying he had close ties with Facebook who gave him permissions for his apps.
Facebook denies this counts as a breach though vice-president Paul Grewal said: “Protecting people’s information is at the heart of everything we do, and we require the same from people who operate apps on Facebook. If these reports are true, it’s a serious abuse of our rules.”
Graphic showing main players in Cambridge Analytica case
The amount of data CA bought access to got so big that it automatically shut down app’s ability to grab profiles temporarily-but Kogan told someone he talked with an engineer who lifted limits quickly so work went back on within days.
Within months Kogan & CA had built database holding millions US voters plus an algorithm scanning them identifying likely politics & personalities allowing tailored messaging targeting specific individuals-a tactic known as micro-targeting in politics.
On Friday FB announced suspending CA + Kogan pending investigation over misuse connected with project-related user info harvesting claims going forward until clarified fully.
FB says harvesting tens millions profiles via GSR + CA isn’t technically breach since:
They claim Kogan got access legitimately following proper steps but later broke rules passing data onto third parties anyway.”
Source link
