Allowing A Debate on Political Ads Enables Silicon Valley’s Worst Behavior
“The saddest fact,” lamented Nikolai Bordyuzha, a Russian four star general at the 2014 Moscow Security Conference, “is that, in an information war, the one who tells the truth always loses. He is limited by truth, while the liar can proclaim whatever he wishes.” Luckily for autocrats everywhere, social media companies weren’t listening. Russia took full advantage of recommendation engines that optimized for clicks and likes and were indifferent to truth, earning (as opposed to paying for) views of their propaganda numbering into the billions. Big Tech has responded with vacillation, deflections, and misdirections. No debate better illustrates this than that over paid political media, a shadow of a shadow of the services’ earned media problem that continues to enable anti-democratic information assaults.
In 2016, Russia executed an earned marketing campaign on Facebook to influence the outcome of the US presidential elections. The achievement of the effort would have made Facebook’s top clients’ mouths water: the campaign included four out of the top five most shared, liked, and commented on political posts in the quarter leading up to the election. The top twenty posts had amassed a staggering 8.7M engagements. With only 1 in a 1000 views of a post resulting in an engagement, and, in Facebook’s own words, these posts representing “a tiny fraction of the total”, Facebook’s newsfeed algorithm could have very well delivered Russian propaganda to every single American voter dozens of times over. For however unthinkable would be Facebook CEO Mark Zuckerberg claiming a campaign that earned a client tens of billions of views had no effect on the client’s sales, he’d confidently claim that Russia’s tens of billions of views affecting the election was “crazy”.
On May 10, 2018, Facebook announced the Internet Research Agency, a Kremlin-linked troll farm with hundreds of employees posting daily to Facebook, had purchased ads worth $100,000 over two years. The story took off. Facebook, rather than defending why its top organic influencer is an anti-democratic autocrat, needed only defend a rounding error in its $27B of 2016 revenue. The argument could then shift from how Facebook’s hallmark newsfeed algorithm, the propellent of its astronomical growth, became the world’s foremost purveyor of disinformation to an argument over censorship. Whether Facebook had any intention in moving the narrative to more defensible ground, or even has fully reckoned with its impact in 2016, is unclear, but the benefit is certain as is the lesson to competitors.
Twitter has proven just as fertile ground for staging information operations. A few months after Bordyuzha spoke in Moscow, Russians used Twitter to stoke panic in a small Louisiana town by feigning a terrorist attack. When a full expose on the information attack, its successor attacks, and the tactics deployed made the front page of the New York Times Magazine in 2015, Twitter failed not only to attempt a remedy but even to issue a response. A year of inaction and silence followed. In the months leading up to the 2016 election, a full fifth of political tweets had come from bots. Fake news tweets outnumbered tweets from official political accounts two to one. A third of human users retweeted bots, and a tenth of the most influential accounts were automated. According to the Oxford Internet Institute, “bots reached positions of measurable influence during the 2016 US election.” None of these studies ever implicated Twitter’s political advertising, amounting to under 0.1% of total revenue, but according to Twitter founder and CEO Jack Dorsey, the company “believe[s] political message reach should be earned”, the exact issue the studies exposed.
Silicon Valley has at every turn refused to address the underlying causes of their role in enabling authoritarian messaging to dominate the information environment. Time and again, they’ve preferred motion to progress, whether applying the same labels to autocrats’ mouthpieces as to publicly chartered newsrooms or using statistical slights of hand to exaggerate their accomplishments. The focus on political ads, a source of no real issues, serves only to enable the tech giants to continue the stand by as their algorithms erode democracy.
