Should we be worried about the Cambridge Analytica/Facebook data row?
Dr Julia Powles, research fellow in technology law and policy at New York University School of Law and Cornell Tech, says YES.
Rogue data science outfit Cambridge Analytica wrongfully acquiring 50m Facebook profiles is a huge deal. This is true whether or not it influenced Brexit and the US election. It’s especially true given that Cambridge Analytica’s operations are business-as-usual for Facebook, a stumbling war chest of vulnerabilities on one quarter of the planet’s population.
Facebook has played the victim, saying researcher Aleksandr Kogan lied when he scraped personal information from 50m “friends” of 270,000 users of his personality research app “thisismydigitalife” and passed them to Cambridge Analytica, which generated psychometric profiles.
At the time, Facebook demanded data deletion. But it did nothing about the models built or insights gained, allowing them to be weaponised in political sociodrama. Despite his seemingly egregious breach of research ethics, Kogan kept his university post.
The scandal plays to deep anxieties about the consolidation and misuse of our most intimate information, demanding a reaching, radical response.
David Sumpter, author of the forthcoming book Outnumbered: From Facebook and Google to Fake News and Filter-bubbles – The Algorithms That Control Our Lives, says NO.
The idea that Cambridge Analytica could use this data to conduct “psychological warfare” is, in my view, far-fetched.
Facebook “likes” can help reveal unexpected relationships between our political sympathies and our favourite films and music – for example, Democrats tend to be Harry Potter fans and Republicans enjoy camping. But likes aren’t accurate enough to reliably determine our personality. If we put two peoples’ Facebook profiles into a personality algorithm and ask it to tell us which of them is most neurotic, the probability the algorithm gets it right is about 60 per cent. Not very good when we remember that a random choice has 50 per cent accuracy.
While scientists have found statistical relationships between our state of mind and what we post on social media, these are not yet strong enough to target our personalities and manipulate us. We are a bit too smart for that. So let’s not get too worried.