Academic who collected 50 million Facebook profiles: ‘We thought we were doing something normal’

The academic at the heart of the Facebook data scandal has said he is being used as a scapegoat by the US tech giant. Aleksandr Kogan, a researcher at Cambridge University, collected a dataset of some tens of millions of Facebook users four years ago using a personality quiz app. Kogan later passed this information on to voter-profiling firm Cambridge Analytica, which claimed (but now denies) that it used the data to craft political ads for President Trump’s 2016 election.

Kogan’s comments, made to BBC Radio 4’s Today program, describe an environment of permissive data-gathering and lax privacy policies. “We thought we were acting perfectly appropriately. We thought we were doing something that was really normal,” Kogan said. “My view is that I’m being basically used as a scapegoat by both Facebook and Cambridge Analytica.”

Kogan said that he was assured by Cambridge Analytica that “thousands and maybe tens of thousands of apps were doing the exact same thing” and that “this was a pretty normal use case of Facebook data.” (Kogan says he handed over information on 30 million users to Cambridge Analytica, although reports from The New York Times and others set this number at 50 million when including profiles with scant information.)

Facebook says Kogan violated the company’s data policies by using the information he collected for commercial purposes. Between 2007 and 2014, the company gave developers access to its social graph — the map of users’ networks of friends, interests, and likes. But multiple reports say the social network did very little to police this sort of activity, only asking third-parties to sign minimal agreements, and only investigating misuses after they were reported to the company.

One former Facebook employee told The Wall Street Journal that “the main enforcement mechanism was call [developers] and yell at them.” An anonymous app builder who collected information from users over this period told Business Insider he remembers thinking to himself: “Fuck, people will literally give away everything for nothing.”

Industry insiders say the collection and misuse of this data was and is common. But the Cambridge Analytica scandal, which broke this weekend thanks to the testimony of a former employee, Chris Wylie, has shone a spotlight on these practices. This is at least in part because of Cambridge Analytica’s involvement in the 2016 US election, where it served as the data operations team for the Trump campaign. Cambridge Analytica has been the subject of parallel investigations, with an undercover reporter filming CEO Alexander Nix boasting about using bribes and sex workers to entrap politicians. Nix was suspended from the company earlier this week.

Read more: How to use Facebook without giving away your data

Whether or not the Facebook data helped swing the election for Trump is unclear, although experts are skeptical. “I think Cambridge Analytica is a better marketing company than a targeting company,” one academic studying political microtargeting told The Verge this week.

Speaking to Radio 4, Kogan echoed these sentiments, saying Cambridge Analytica had tried to sell “magic” but that its efforts “could have only hurt [Trump’s] campaign.”

“The accuracy of this data has been extremely exaggerated,” said Kogan. “In practice my best guess is that we were six times more likely to get everything wrong about a person as we were to get everything right about a person. I personally don’t think micro-targeting is an effective way to use such datasets.”

But even if microtargeting is not effective, it’s clear that these reports have brought Facebook’s reputation to a new low. The #deletefacebook campaign is in full swing, and even high profile figures like Brian Acton, co-founder of WhatsApp, which Facebook bought in 2014, have joined in. The social network is reportedly under investigation from the FTC, from New York and Massachusetts’ attorneys general, from Canada’s privacy watchdog, and its CEO Mark Zuckerberg has been summoned by UK MPs to answer questions before parliament. Zuckerberg has been notably absent this week, and has yet to make a public statement on the scandal. He’s expected to do so before Thursday evening.

The Cambridge Analytica scandal

What happened 7
Facebook reacts 8
The world reacts 13
The apology tour 6
Zuckerberg testifies 14
What you can do 5

Comments

They were doing something normal. Obama administration officials have been quoted on record that they had basically unrestricted access to FB for targeting during his re-election. They were hailed as heros by the tech world for creating such an amazing system that was so far ahead of Romney’s. Now it turns out Trump did it and the world is on fire? Please, I’m no Trump fan but this is hilarious watching everyone lose their mind over something FB and companies have been doing for years.

Public awareness. That’s the difference here.

That overwhelming majority of FB users are ignorant as to what FB does with their data. Hellz, they don’t even consider it.

They were upfront about who they were and what they were doing. This company was not and collected data they should not have collected under Facebooks TOS. And then they gave it to a super shady political research group that has run rumor campaigns to discredit elections.

I wonder myself if, had Bernie won, we would’ve even cared at all. How many times do we need to be reminded that selling our data is and has always been FB’s bread and butter?

https://slate.com/technology/2018/03/facebooks-data-practices-were-letting-down-users-years-before-cambridge-analytica.html; this article explains how FB advertised their own consultation services to political parties. Again, is this even surprising to anyone, really? Free products aren’t free. This sudden uproar seems ever so slightly hypocritical; are we still trying to rationalize Trump’s win?

Again. Using targeted advertising like FB is advertising is not the problem here. Lying about why you are collecting data, then taking that data off to a 3rd party against the terms and conditions of the use of that data is. One is playing by the rules and above board. The other is not and illegal.

Using Facebook’s targeted ad platform is not the issue here. One entity bulk collecting the data, then passing to a political analytics team to process for a candidate is a different thing entirely.

Still, the problem here, isn’t exactly a particular candidate or campaign. The problem here is that Facebook doesn’t enforce its rules. They let people collect the data in a way that leaves them no control over it once it has been collected. I’m sure a lot of private companies are doing exactly the same thing.

But if, after getting reprimanded for what was just uncovered, FB decides to do exactly what Analytica did, but in-house, would there be a difference?

Does it actually make a difference to the casual user that it was External Party XYZ and not FB themselves picking at the data to influence behavior? Especially when doing so was already at the very core of their business (whether intended or not)…

Again, they had the same level of control during Obama era. Except they were transparent about its use, whereas in this case, the information was leaked via a whistleblower hence why the tone is drastically different. But in effect, weren’t they doing the same thing?

Yes, it does matter. One was legal and following the rules for how FB data can be used. One was not. One was targeted advertising, more about getting out the vote following well published capabilities, the other we don’t really know what kind of techniques were used or what kind of manipulation was attempted because it happened inside a 3rd party’s servers.

Yes, it does matter if a party says they are collecting data for academic research and then turning it over to a political analytics company. That is not at all the same things as buying ads and using the same tools everyone else has available. Your attempt to equate the two is weak.

Just another reason to never have a FB account.
The main problem here seems to be that two "cultures" that have the skewed world view that everyone is "good" setup a system that collected the data for "research". "Research" is in their world view always "GOOD" and the data they collect is always only going to be used for "GOOD".
Now in FB’s case, "GOOD" is defined as making the Zuck and friends rich so they can use the money to do "GOOD".
In the researchers case "GOOD" means they may someday be able to pre-identify a serial killer (or terrorist, or pedophile, or , or , or christian baker,or Trump) and stop them before they do whatever it is the researcher thinks society thinks should be stopped. Bear in mind, the researchers are almost universally of a Marxists mentality so anything "conservative" is bad and since they never associate with anyone that doesn’t agree with them, they think they are absolutely justified in trying to root out conservatism.
This is the real reason the researchers and the techies are so upset. Their very first premise that all science and research is "GOOD" and will only be used for the things they think are "GOOD" was violated by someone supporting everything (conservatism) they think is "BAD".
But, money and good intentions will win out and the data collection will continue and the misuse of the data will also continue. So, your only defense is to not have a FB account and use every tool available to limit the information you give to any other "social media". Turn off photo and location sharing. If a field is optional, opt not to fill it in. If you MUST fill it in and you cannot see any good reason for the request but you must make use of the service, such as buying something online, put in false information. The worst they can do is cancel your account at which point, you use a VPN to create a new fake account after wiping out the cookies they planted the first time.

Deleted it like 5 years ago.

Man I’m so ahead of the curve it’s not funny.

Academic that does not understand Informed Consent, thinks normalcy equates to ethical.

View All Comments
Back to top ↑