Rest easy, everyone. A year after President Trump’s inauguration, Twitter claims it has solved the problem of Russian disinformation that plagued the 2016 election by playing — and according to Twitter executives, winning — this round of the game of Whack-a-Troll.
In a Friday afternoon post updating users on its review of the 2016 election, Twitter touted its removal of 3,814 accounts and associated posts from the infamous St. Petersburg troll factory known as the Internet Research Agency.
Facebook has similarly issued several statements this month about planned changes to address disinformation on its platform. Both companies have employed strong language to condemn the use of their platforms for spreading lies; Samidh Chakrabarti, the head of Facebook’s civic engagement team, admitted that Facebook was “far too slow to recognize how bad actors were abusing our platform.”
It’s refreshing to see that these companies have finally realized their culpability in the spread of online disinformation. But their solutions to the phenomenon overlook homegrown purveyors of it. The companies have misplaced their focus in the fight, concentrating too heavily on removing Russian content while ignoring the problematic articles and posts created and shared by American outlets and users.
In its statement, Twitter asserted that it is committed to identifying and removing more bots in the future. But it also added a few important caveats: “automated election-related content associated with Russian signals represented a very small fraction of the overall activity on Twitter in the ten-week period preceding the 2016 election,” and of tweets from Internet Research Agency-linked accounts, only 8.4 percent were “election-related.”
What Twitter does not address is that the trolls will keep popping up. The Kremlin is unlikely to have concentrated all of its information warriors in one building. And according to researchers, up to 15 percent of all active Twitter accounts might be bots, Russian or otherwise. Twitter has a great deal more work to do on the automation front, but Whack-a-Troll is not where its disinformation problem ends.
Twitter does not seem to comprehend that non-automated content, including from American accounts, is a huge part of the disinformation ecosystem. Its amplification of neo-Nazi rhetoric, abusive content and false or misleading stories spread by accounts with huge numbers of followers all affected rhetoric surrounding the 2016 election. Other than supporting a few small media literacy programs, Twitter doesn’t seem to have a plan for eliminating the problematic content that originates within America’s own borders.
Facebook, which played its own round of Whack-a-Troll last year, scrubbing its platform of Internet Research Agency-created pages and accounts, doubled down on efforts to abdicate responsibility for defining high-quality content. On Jan. 19, Facebook announced that it will poll users to determine which outlets they trust and with which they are most familiar, giving those outlets priority in the News Feed.
These polls seem likely to reinforce existing biases, privilege well-known media giants and leave small outfits in the dust. Local news will appear “in a dedicated section,” confined to Facebook purgatory, along with the “Pages Feed” and other content streams that are not automatically served to users. This is a critical mistake. After a year in which the United States witnessed a 37 percent decline in trust in institutions, the country’s dwindling local newsrooms are more important than ever, interpreting and contextualizing events from Congress to the City Council for their readers in a way that larger news organizations struggled to do. Facebook should place local news in a prominent location, prioritize content from high-performing smaller outlets, perhaps through a competitive application system, and make serious investments in local journalism.
Facebook will also continue to rely on users to report false stories, which it will work with third-party fact-checkers to confirm and label. It claims this process stops the spread of untruths by 80 percent. Conveniently, Facebook ignores decades of research demonstrating that those who have consumed untruths are unlikely to buy into the corrected version.
Like Twitter, Facebook is doing little to address problematic narratives that may not be patently false. After all, most of the infamous Russian election ads purchased on Facebook in 2016 were not “fake.” They were highly inflammatory messages aimed at the specific populations with whom they would most resonate. How might a concerned citizen or outside fact-checker respond to an ad comparing Hillary Clinton to Satan, or to a meme arguing that blue lives matter more than black lives?
For both platforms, and for the internet as a whole, this is admittedly a hard question. But it’s not one without an answer. As private entities, these companies have the right — and the obligation — to update their terms of service to reflect the realities of the disinformation era, defining in plain English what content is permissible on their platforms and actively enforcing those definitions. Yes, it would be costly, but you can’t put a price tag on democracy.
By narrowly focusing on eliminating Russian accounts and posts, relying on users to determine which content is trustworthy and blindly believing that fact-checking will improve the rapidly worsening level of civil discourse in America, social media companies are relinquishing their role as today’s most powerful gatekeepers of information. If these tech giants want to contribute to democracy instead of help to tear it down, they need to recognize that homegrown threats to civil discourse exist among the very users to whom they are bequeathing more responsibility. In a world where every person with a smartphone is a citizen journalist, a threat to democracy is a threat to democracy, no matter the post’s country of origin.
Nina Jankowicz is a writer and analyst focusing on disinformation.