Is it too late for Silicon Valley to avert election chaos?

A man in line for a Trump rally in Wilkes-Barre, Pa., in August holds aloft a symbol for QAnon.
A man in line for a Trump rally in Wilkes-Barre, Pa., in August holds a symbol for QAnon, a far-right conspiracy theory that Facebook and Twitter have promised to help curb.
(Associated Press)

There is nothing subtle about the script President Trump is preparing for election day, as he and his supporters — fearing defeat — blitz social media claiming that the spike in what are legitimate mail-in ballots is in fact part of a plot to steal the presidency from him.

Yet the script for social-media executives, who are hoping to avert the chaos Trump’s move would cause if he loses, is much harder to follow. They are still scrambling to write it. And that’s a problem, critics contend.

Two giants took additional actions this week to try to reassure the critics, election officials and voters. On Friday, Twitter announced new curbs on what can be tweeted and how fast it can spread on the platform, two days after Facebook also unveiled new restrictions.

Yet as Nov. 3 nears and early voting is already underway, disinformation researchers and voting rights activists remain unnerved by the timidity of the media platforms in confronting a deluge of Trump-instigated falsehoods and conspiracies. They warn the companies are inviting an election-week meltdown.


“We are at a critical moment, things are dialing up, and we are not seeing a level of urgency from the platforms,” said Andre Banks, whose firm A/B Partners launched a campaign to combat online misinformation, targeted at Black and Latino voters. “We want to see they are as worried as everyone else in America is.”

The groups are sounding the alarm because the platforms’ stepped-up actions, such as dismantling foreign bot networks, adding fact-checks to election-related tweets and posts, and limiting political ads, are failing to stymie the contagion of false claims and conspiracy theories around voting. The companies are navigating volatile territory, facing threats of retaliation from Trump and Republicans in Congress.

Twitter said Friday that misleading claims from politicians will appear behind warning screens that users must click through to see the post. The platform also will make retweeting others’ posts more cumbersome, to slow or avert the viral spread of misleading information.

Facebook officials have gamed out more than 70 scenarios involving misuse of their platform as election results come in, they’ve said. The company added new rules on Wednesday, including extending a ban on political advertising beyond election day, adding fact-checks to posts by politicians prematurely claiming victory, and banning content that “seeks to intimidate voters.”

“We’ve known for a long time that the 2020 election in the US would be unlike any other,” said a blog post from Guy Rosen, a Facebook vice president. “We’ve been preparing for this.”


Critics were underwhelmed. Sen. Elizabeth Warren (D-Mass.) branded the new Facebook restrictions “performative.”

The companies have grown considerably more savvy to the threat of disinformation since 2016, when they did not comprehend the scope and impact of online interference from abroad until long after that year’s election. They are aggressively removing content placed by foreign operatives.

But the bigger threat in this cycle are domestic provocateurs, including the one in the White House.

“We are much more concerned this cycle about domestic disinformation, and it is pretty evident the platforms are failing on that front,” said Jesse Lehrich, co-founder of the advocacy group Accountable Tech and a former aide to Hillary Clinton and President Obama.

The fact-checks regularly come too late or not at all, and they are often ineffective. The companies are loath to silence the accounts of repeat violators of their policies when those offenders include prominent lawmakers and the leader of the free world. Their algorithms routinely drive users to incendiary and misleading information, yet the platforms resist calls for more transparency around them.


While they have robustly fought disinformation about the coronavirus, they have been reluctant to use the same aggressive tactics with unsubstantiated claims of election-rigging. As the pandemic spread, “they acted immediately and were willing to out falsehoods and delete them,” said Ann Ravel, the former chair of the Federal Election Commission. “They are not as willing to do that with issues related to the election.”

The nightmare election-day scenario is obvious by now, disinformation scholars say. Mail-in ballots lean heavily Democratic, polls and voter requests suggest, and in some states they aren’t counted until after the polls close. Trump and the networks of right-wing activists supporting him have already conditioned vast numbers of his backers to believe that, if he falls behind after the in-person votes are tallied, the reason will be fraudulent mail-in ballots.

This “pre-framing,” said Kate Starbird, a data scientist at the University of Washington, has been so persistent and pervasive that the measures companies are finally taking to confront it are proving ineffectual.

“It is hard to put a wrench in this wheel,” said Starbird, who helps lead the Election Integrity Partnership, a group of prominent scholars working to detect and mitigate online misinformation and disinformation around the election.

“Each time a voter sees a piece of evidence taken out of context, it just reinforces these messages that a lot of audiences have been exposed to for some time, that they can’t trust the election results,” Starbird said.

One obvious example is the far-right conspiracy theory QAnon, which Facebook only banned this week. Its proponents had been operating on the platform for years, spinning disinformation to tens of thousands of followers. The groups amplified, among other falsehoods, the debunked theory — spread with the help of the Trump campaign — that Joe Biden was wearing an earpiece during his debate with the president on Sept. 29.


The resonance of such disinformation was clear at a small grocery in rural Brooklyn, Iowa, where five elderly women sat around a table talking politics the day after the debate. The phantom earpiece was a hot topic.

Dorothy McConnell, 83, a retired secretary and school aide, insisted the former vice president was wearing one, so advisors could tell him what to say on stage. “Biden’s not all there,” the Trump supporter said.

False reports of mail ballots getting dumped in ditches are proving just as durable. A case in Wisconsin that Trump and his advisors keep citing to disparage mail voting did not involve any state ballots getting tossed, election officials reported. Yet more than 65,000 Twitter users, each with at least 10,000 followers, tweeted about the incident over two days in late September, according to the Election Integrity Partnership.

By early this week, Facebook had banned ads that brand mail-in ballots as corrupt or fraudulent, or use “isolated incidents of voter fraud to delegitimize the result of an election.” But the company remains cagey about what further steps it will take in the critical weeks ahead.

Facebook and the other companies are facing stiff political headwinds. Trump and many Republicans complain of censorship in the companies’ efforts. As the election approaches, some in Congress, with Trump’s support, are again moving to strip “Section 230” legal protections that shield the social media companies from lawsuits when they fact-check or take down misleading content.

“Facebook is actively targeting ads by conservative groups ahead of the election, either removing the ads completely, or adding their own disclosure if they claim it didn’t pass their fact-check system,” Sen. Rick Scott (R-Fla.) said last week, just before the Senate Commerce Committee summoned the CEOs of Facebook, Twitter and Google to appear before it Oct. 28.


He complained that the companies rely on nonpartisan fact-check organizations that he sees as biased against Republicans. “Liberal media outlets getting to say what ads are on a social media site? Gimme a break,” Scott said.

Emma Llansó, director of the Free Expression Project at the Center for Democracy and Technology, called the GOP push “a direct attack on the ability of services to do things like fact-checking. … We have never seen anything like this — certainly not right before an election.”

A Democrat on the Commerce Committee, Sen. Brian Schatz of Hawaii, was more blunt. He accused Scott, fellow committee member Sen. Ted Cruz (R-Texas) and other Republicans of clearing the way for Trump to use social media to declare victory before all the ballots are tallied, and “declare the continued counting of lawful votes as a coup, and that people should rise up.”

Schatz said that Republicans’ renewed push to repeal the legal protections essential to the companies’ business model “appears to me like an attempt to work the refs coming up to an election.”

Times staff writer Noah Bierman in Brooklyn, Iowa, contributed to this report.