US Election Article Bar
The way Sir Nick Clegg tells it, one cannot help picturing a giant red button in Mark Zuckerberg’s office.
“There are some break-glass options available to us if there really is an extremely chaotic and, worse still, violent set of circumstances,” said the former deputy prime minister, now Facebook’s chief political fixer, in an interview with the Financial Times last month.
Like most events in 2020, it would have seemed an extraordinary statement only a year ago. Facebook, Sir Nick gravely explained, was preparing to use "exceptional measures" created to limit mass violence in countries such as Sri Lanka and Myanmar for the same purpose in the US presidential election.
Yet after a summer of protests, riots, police brutality, terror plots and militia threats, it did not seem so strange that Sir Nick should echo his boss Zuckerberg’s repeated warnings that the election could descend into "civil unrest".
"Companies like ours need to go well beyond what we’ve done before," he told investors last week. "[This] will certainly be a test for Facebook."
Now, with mere days to go until the polls close, we have a pretty good idea of what that red button does – and of just what Zuckerberg and Sir Nick are so afraid of. Alongside counterparts at other tech firms such as Twitter, Facebook is scrambling not only to shut down incitement and disinformation reactively but to change the fundamental building blocks of its apps.
Election violence ‘could happen here’
"There’s long been a sense that the challenges that are facing so many countries, in terms of unrest, threats to democracy and human rights, are things that happen ‘over there’," says Chloe Poynton, a former UN official and human rights consultant commissioned by Facebook to probe its role in violent mob attacks on Sri Lankan Muslims in 2018.
Poynton’s report found that Facebook may have helped enable the violence by letting viral hate speech spread unchecked (it only had two moderators who spoke the local language), and led the company to apologise this year.
While she is not authorised to talk about that work, she warns that social networks have been "weaponised" by extremists across the world much as the Rwandan genocidaires weaponised radio.
"There’s a real sense of American exceptionalism, a real belief that our institutions are stronger than any other institutions," she says. "And I don’t see why that is necessarily the case. I think that what has happened in other countries could easily happen here in the US."
What would that look like? One clue comes from a series of crisis simulation wargames run by the online security firm Cybereason, exploring how well-organised hackers might crash an election by flooding it with cheap disinformation – what experts call a "perception hack".
"We sent suspicious envelopes of white powder to polling places to get them shut down," says Maggie MacAlpine, describing her role in one such game last month. As the non-partisan co-organiser of an annual voting machine challenge at the hacker conference Defcon, she was a good fit for the "red team" of fictional election wreckers.
She and her team-mates deployed fake tweets calling on citizens to take up arms against illegal immigrants trying to vote; photos of long lines from previous elections; fake rumours of "gun-toting militias" intimidating voters; and a computer-generated "deepfake" video of one candidate giving a concession speech.
The Telegraph played the same role in another Cybereason game in February, spreading fake videos of malfunctioning voting machines, calling in bomb threats, sending official-looking emails to all polling stations telling them to turn some voters away, and creating a deepfake of ballot boxes being removed."
The real danger is after election day
These scenarios are fantasies, but most of the individual methods are possible. Iran already appears to have attempted a perception hack, sending emails to voters in Florida purporting to be from right-wing extremists threatening violence if they do not vote for Trump, with the probably intention of fomenting distrust in the vote.
And while a single, coordinated operation of the scale of Cybereason’s game is unlikely, this election will be particularly vulnerable to destabilisation. What makes poll-watchers break out in a cold sweat is a period that some have called "the interregnum": the wide gap between polling day and the counting of results that will open this year due to mass mail-in voting.
US Election 2020 – when is the winner declared?
According to MacAlpine, this is the first year that voting methods have become a partisan issue, with Democrats favouring postal votes and Republicans preferring to go in person. That could create a "red shift" or "blue shift" effect in which projected results lean strongly in one direction and then "swing wildly" back again.
"What we’re afraid of is if one of the candidates declares victory prematurely based on very early results, and then galvanises their followers to take to the streets because ‘the election is being stolen’," she says.
Indeed, numerous former US intelligence agents, military bosses and congressmen have told the Telegraph that they fear President Trump may abuse the power of the White House, and his bully pulpit, to stay in charge during the gap.
We strongly believe that everyone who is eligible should be empowered to #vote.
And we know that people on Twitter are seeking resources to safely participate in the US elections, which is why today Twitter is encouraging people in the US to vote early. pic.twitter.com/qw8XSObbxE
— TwitterGov (@TwitterGov) October 22, 2020
Extremists, armed militias and anti-fascist groups could take matters into their own hands, and crowds of protesters could form unpredictably. Wild rumours, fabrications and exaggerations about opponents’ behaviour might escalate into calls to action, and then into calls for violence.
Which is, in theory, where tech firms come in. From Facebook’s centralised "elections operation centre", in constant communication with other tech firms and government agencies (head of cybersecurity Nathaniel Gleicher joked on Friday that he calls his counterpart at Twitter more often than the man’s mother does), the company will watch trends building on its networks via custom data tools.
In the worst scenarios, it could suddenly slow the spread of all viral content, turn up the sensitivity of its hate speech hunting algorithms, or shunt certain kinds of content out of users’ news feeds.
Twitter almost certainly has similar weapons, having barred its users from sharing a controversial news story about Joe Biden and blocked some users from tweeting at all during a disastrous cyberattack in July.
Tech giants fear their own systems
Yet these reactive measures might not be enough. Content that goes viral tends to do so within the first few hours of posting, or even the first 30 minutes. And so both Facebook and Twitter have also begun to rethink foundational elements of their services, reversing their traditional obsession with making it faster and easier for users to share and consume.
Instead, Twitter has made it harder to retweet and is adding warning labels and pop-ups to swathes of content, from false information about voting procedures to articles that users try to retweet without having read for themselves. It will also stop pushing tweets from people that its users do not follow into their feeds.
Facebook, meanwhile, has removed a key news tab from Instagram and banned political advertising indefinitely after the election. Notably, it has quietly removed all political Facebook groups from its algorithmic recommendations, despite a massive push over the past few years to get more people into groups.
The company has privately ruled out shutting the whole service down, which makes some sense: without social media, voters would be left without any ability to share and document actual instances of violence and malpractice, so there is no easy way out of the trap.
Even so, both Facebook and Twitter’s moves are a surprising departure from their gunslinging past, when growth was king and user engagement almost always beat safety. Longtime critics describe it as an effective admission that social networks’ core features are, and have always been, dangerous.
The companies’ actions are also likely to attract fury from some elected politicians and many voters, who will accuse them of suppressing crucial information about genuine irregularities or scandals.
"The planning appears to be occurring far too close to election day, when this threat has been discussed constantly for over four years," says Lisa Kaplan, head of the disinformation tracking firm Alethea Group.
"Until there are systemic changes, voters need to realise that no one is coming to save them from disinformation, and they must rely upon themselves."
Свежие комментарии