A referendum is not a beta test
9 April 2018 | 0
In my last post I wrote about the firefighting measures Facebook has promised in the wake of the Cambridge Analytica scandal where 87 million account holders had their data used without consent to construct psychometric profiles. Users fitting these profiles were then targetted with advertising dressed up as news stories favouring the Trump campaign.
So far the CA scandal has shown us how easy it is to a) harvest data b) use said information to generate fake content c) avoid censure for breaking Facebook’s terms of service, and d) create a sustainable fake news economy by reinforcing people’s confirmation bias, converting clicks into votes.
It’s still hard to know how effective this strategy actually was. My guess is partisan politics, gerrymandering, a candidate with a populist message and a decision not to effectively campaign in three states doubtless played a bigger part in securing a win for the Trump campaign.
Facebook’s sin, according to CEO Mark Zuckerberg, is that the company trusted the market to self-regulate and users to go deeper than reading clickbaity headlines. Facebook trusted users to know better, users trusted Facebook to be better, bad actors exploited the space between and now we are where we are.
Last Friday the fightback began in earnest when Facebook began rolling out of a series of new privacy measures, including a notice of whether your data was compromised by CA.
The announcement also showed a new approach to political and so-called “issue ads” on Facebook and Instagram.
Under measures being trialled in the US, political advertisements will have to be labelled as such, along with information on who paid for them. Facebook will also moderate who can run such ads and demand information on identity and location. If you can’t prove you’re not a Russian troll farm, you won’t be able to get on the platform. In theory.
I’m excited about these measures. It can provide a fascinating window on the world of political campaign funding.
Unfortunately we won’t be seeing the benefit of Facebook’s newfound sense of responsibility for some time. Given next month’s plebiscite on the 8th Amendment and allegations of foreign campaign funding from both sides, being able to strip away at least one vector of misinformation could have a significant effect on voter behaviour.
It’s disappointing that Irish voters won’t have these rudimentary protections in place by 25 May when we go to the polls, but doubtless Facebook has its eyes on the November US mid-term elections. After all, a referendum is not a beta test.