I’ve been keeping track of how social media companies have been preparing for the 2020 US Presidential Election, and wondering if we’ve learnt anything at all from 2016, considering they’ve had FOUR YEARS to fix problems (like that’s a seriously looooooong time, although I will cut some slack for the covid, even if that misinformation also seems to heighten the urgency). It wasn’t just about privacy and data, it’s also about hate and bigotry, and it’s also about voter suppression.
Rather than choose to deal with these problems on social media tools holistically, these companies have chosen to deal with the US election specifically. Twitter announced an expanded policy in October, just a month before the election. Facebook followed suit, announcing a Voting Information Center as well for Facebook and Instagram. Google already suspended political ads, and Youtube has notifications in search and directs users below videos to Google’s election page with Associated Press sources. TikTok has no political ads and is often underestimated for how it will influence the election. But it specified that it will reduce the visibility and social spread of any premature claims to victory, and direct users to AP results on its election guide. Pinterest is pretty vigorous, but chat companies seem to lack policies (Telegram, Facebook-owned Whatsapp has limited the number of people a message can be forwarded to), as do game-streaming companies (Twitch, Discord).
(Interestingly, it’s only when Western democracies are in trouble that any lawmaker cares about a global company. Meanwhile, there’s an election in Myanmar coming up, where there has previously been social media disinformation with effects on human life itself.)
As the results of the election continue to roll in, it has been interesting to see how Twitter, in particular, has reacted—in line with their Civic Integrity Policy And they’ve put warnings on several of Donald Trump’s tweets, not least of which includes this one on Election Night:
Then came the next one:
And by the time these ones on lies about election fraud had been tweeted, Trump’s timeline was more hidden than visible:
Twitter hid plenty of tweets from Georgia’s newly elected QAnon-supporting congresswoman, and one has to wonder if these companies had been more proactive with their election policies earlier in the year, whether she would even have been elected. It also chose to use 7 news outlets as their sources for calling the race. Twitter is well into the process of becoming a publisher and not a platform.
I recently noticed that it was splitting breaking news into the headline and summary, original tweet, a “Read More” consisting of tweets with links to news articles, “Reactions from the White House administration” with an explanation that Mark Meadows is Trump’s Chief of Staff, and “Analysis from reporters”. See the case study of the breaking news that former Department of Home Security official Miles Taylor wrote the anonymous New York Times op-ed and book, A Warning:
On the other hand, Facebook fell a bit short. Along with their election policy, they also decided not to accept new political ads in the week before Election Day, and ban all political ads after voting closes indefinitely. But older ads could still circulate and be retargeted, and Facebook doesn’t actually fact check any ads anyway. In fact, the policy ran into problems—some ads were being blocked despite having approval—and Joe Biden’s campaign wasn’t happy either. (Update: Facebook also tested a new AI-powered misinformation detector in the months before the election, though it is unclear if it was actually used.)
And previously, Facebook had tweaked its algorithms so as to reduce the news people would see in their News Feeds, citing that it would “bring people closer together”. In fact, insiders say that Facebook had appeased Republicans and it ran experiments, knowing it would have a disparate effect on news outlets depending if they were conservative or liberal. And the media already has a conservative bias, because journalism isn’t actually the tall order of the day. (The Daily Wire‘s Ben Shapiro has hung out with Mark Zuckerberg, and there are other Republican operatives in the Facebook Washington office).
In fact, TikTok and Facebook both took down hashtags related to false claims of victory and violence. But Youtube refuses to take down videos despite them apparently violating their own policies, though it’s also the title, thumbnail and description that also affects how viewers interpret the video itself.
Youtube and Facebook did take down Steve Bannon’s livestream in which he called for the beheading of two government officials including Dr Anthony Fauci and Chris Wray, but it took Facebook 11 hours to do so, and he was back the next day with 6.9k concurrent viewers (though it didn’t appear on the Facebook Live popular page). Twitter suspended his account permanently. (Update: Facebook took down a group of pages affiliated with Bannon that had pushed disinformation about the election on Monday 9 November—with a total following of something like 2.5 million people.)
Facebook finally shut down the 350,000 strong public group called “Stop the Steal”, though it would seem others are also proliferating. In fact, thanks to this Buzzfeed scoop, we now know Facebook has an internal metric for “Violence and Incitement Trends”. The fact that it exists, let alone that it is rising during Election Week, is frankly alarming. Meanwhile, two fake Trump posts were the most popular posts on Facebook:
Instagram removed recent posts for hashtags to slow the proliferation of false information (for some users—it depends on updating the app as it’s not algorithmic change), though I somewhat doubt its efficacy. If anything, it forces people onto Stories, which are far less moderated, since they last a mere 24 hours, which would require far too many human moderators for such quick turnaround. Instagram does flag some disinformation by greying out the photo in the home and profile feeds, putting an icon of an eye with a warning on top (both covid- and election-related). But I remain unconvinced anyone is taking it seriously enough. My Instagram feeds were full of VOTE lines under each post (as well as in Stories), linked to the voting centre:
The thing is, social media companies alone are not to blame. Fox News appears to be lauded for finally calling out Trump’s lies for what they are, as does much of the rest of the mainstream media, despite the fact that rampant lies remain uncontained on other right-wing media outlets. And left-wing media outlets have likewise failed to report effectively on Trump’s lies since the 2016 election cycle, though they seem to have grown a spine in the past week by cutting away from Trump’s press conference lies. These lies don’t live on social media tools exclusively; rather, they are part of a whole media ecosystem and the mainstream media fails to see themselves for this:
And, by waiting for a 0.5% difference before calling Pennsylvania for Biden, Associated Press, and by extension social media companies, are leaving wide open a vacuum for disinformation and for Trump to tweet absolute lies. I understand the natural impulse to be cautious after the 2016 election, and after the earlier-than-expected call of Arizona for Biden, but there IS such a thing as being too cautious, especially when it allows disinformation to spread, particularly on social media. (Update: not to mention that pretty much every other major media outlet called minutes after AP, which seems … suspicious.)
There are no two ways about it, social media is now the mainstream media. I myself have seen SkyNews Australia on the “breaking news” section of my Youtube home feed, despite having never watched a video published by them before, and that’s all part of the machine.
So is it too little, too late? Social media companies will take time to get their policies into place, though they undoubtedly cannot do this alone. For instance, the URL to the Hunter Biden story in the New York Post was blocked on Twitter and their account was frozen, though CEO Jack Dorsey later retracted this decision and made platform-wide changes to the algorithm instead. Global governance is necessary to change this profit model and make it work for the public good, rather than for the rich few.
But the future of democracy relies on us getting this right. We’ve seen this play out in America, the world’s biggest power. Imagine what could happen in any other country.