Tech platforms are racing to combat misinformation before elections. Is it too little, too late?

April 11, 2019 and
Category: Fact-Checking

Factually is a newsletter about fact-checking and accountability journalism, from Poynter’s International Fact-Checking Network & the American Press Institute’s Accountability Project. Sign up here.

Facebook plays election whack-a-mole

This month, elections kick off in at least five countries around the world. As a result, Facebook and other tech platforms have been doing some last-minute preparation to try to protect their platforms from misinformation.

In India, where voters have been stricken with misinformation in the lead-up to the start of the general election this week, WhatsApp announced a new research project aimed at learning how rumors spread on the messaging app. That move, which launched just last week and was initially billed as a verification project (it isn’t), came more than a year after fact-checkers started debunking hoaxes on their own institutional accounts.

Also in India, Facebook added in February five new fact-checking partners to find and debunk false stories, images and videos on the platform. It has removed hundreds of accounts and pages for “coordinated inauthentic behavior.” And, according to the Hindustan Times, the company also has 40 teams of 30,000 people around the world looking for potentially false or problematic content that could affect elections like the one in India.

In Indonesia, where for-hire “buzzers” spread deliberately false information about both presidential candidates on social media, fact-checkers have been busy debunking political hoaxes ahead of the April 17 election, France 24 reported. In March, Facebook restricted foreign advertising and shut down several accounts and pages spreading hate speech and misinformation.

These attempts vary in size and importance, and all reveal that Facebook seems to take election misinformation seriously. But is it enough to combat the problem at scale and on time?

Save for India, where Facebook reportedly started working on addressing election misinformation 18 months ago, most of the actions Facebook has taken to try to prevent election misinformation have been relatively last-minute.

In Israel, there’s only one fact-checking outlet debunking false political hoaxes on Facebook — and they were brought onto the program just days before Wednesday’s election. Last week, the company announced a similar move in Australia, where elections are set to begin next month. Facebook also announced that it would restrict foreign advertisements.

Giving fact-checkers and new anti-misinformation policies only a few weeks before elections seems like a short time to limit the flow of election misinformation. At the same time, misinformation is a year-round problem with year-round consequences — and not all of them are political.

Take, for example, the recent measles outbreaks around the world, which have run parallel to an influx in anti-vaccine conspiracy theories on platforms like Facebook and YouTube. Or look to the rise of vigilante mob violence resulting from the spread of child kidnapping rumors in France, Mexico and India.

Election misinformation is surely an important problem to address — but it’s just one piece of the pie. Going forward, tech and media journalists would do well to remember that, and hold companies accountable accordingly.

…technology

  • Facebook has announced several big updates on how it addresses misinformation on its platform. Among them: Facebook is now reducing the reach of groups that repeatedly spread misinformation, exploring the use of crowdsourcing as a way to determine which news outlets users trust most and adding new indicators to inform users about the content they’re seeing. In other news, the Associated Press has expanded its participation in the company’s fact-checking project.
  • Last week, WhatsApp announced that it had launched a tip line for verifying rumors during the Indian election. But turns out it’s not for fact-checking misinformation at all — it’s for research.
  • “Welcome to 2019, where it takes just 19 hours for a faked homemade video of Joe Biden to travel from the keyboard of a pseudonymous ‘memesmith’ to the president of the United States,” Charlie Warzel wrote for The New York Times. Here’s what happened.

…politics

  • The Asia Internet Coalition, an industry association whose members include Facebook, Google and Twitter, said Singapore’s proposed law against “fake news” poses “significant risks to freedom of expression and speech.” The Washington Post, similarly, asked in an editorial whether Singapore is fighting “fake news” or free speech. Here’s Daniel’s updated guide to what other countries are doing in the misinformation sphere.
  • The United Kingdom this week announced its plan to create an independent regulator for “online harms.” It must be approved by Parliament. The BBC detailed the proposal, with analysis from Rory Cellan-Jones who said a whole host of questions still need to be answered. Canada’s going there, too.
  • In November, France passed a law that requires tech platforms to publish who has purchased sponsored content or campaign ads and for what price. Now, the French government itself is getting into trouble with that law.

…the future of news

  • A House Judiciary Committee hearing Tuesday in which Facebook and Google officials testified on online hate broke down into partisan sniping; adding to the tension was the fact that a YouTube livestream of the hearing drew racist and anti-Semitic comments.
  • Snopes is launching its first-ever paid membership program. It has also stopped using the Schema.org ClaimReview markup, which Google uses to surface fact-checking articles in search results. Why? “I want them to license the claim review data or prove the rich snippets they create with it are beneficial to the publisher,” Vinny Green, vice president of operations, wrote in a tweet.
  • Our own desire to justify our decisions, and our flexibility in making up our minds, can contribute to our absorption of “fake news,” author and brain scientist Jeff Stibel wrote in USA Today this week.

Each week, we analyze five of the top-performing fact checks on Facebook to see how their reach compared to the hoaxes they debunked. Check Poynter.org on Friday for more on this week’s numbers.

  1. Factcheck.org: “Concocted Claim of McCain’s ‘Treason’” (Fact: 5.5K engagements // Fake: 2.9K engagements)
  2. Maldito Bulo: “No, neither Muslims nor immigrants receive free medicines” (Fact: 3K engagements // Fake: 117K engagements)
  3. Liputan 6: “Are Circulating Photos of Ocean of Humans in Jokowi’s Campaign in Lampung Real?” (Fact: 693 engagements // Fake: 505 engagements)
  4. PolitiFact: “No, gun control regulation in Nazi Germany did not help advance the Holocaust” (Fact: 604 engagements // Fake: 735 engagements)
  5. Africa Check: “Beware false cancer cures from ‘Dr Gupta’ on Facebook” (Fact: 300 engagements // Fake: 15K engagements)

Our choice this week is not a specific fact-check but a collection of them, all on the same topic.

In remarks Tuesday in the Oval Office, U.S. President Donald Trump said the practice of separating migrant children at the southern border took place under the Obama administration. A number of American fact-checkers debunked the claim immediately – and they were able to do so more quickly than usual.

Why? Because this assertion has been checked before.

The New York Times called the claimoutright revisionist history.” The Washington Post’s Fact-Checker noted that it had previously assigned four Pinocchios to the claim. CNN said it is “false and requires context.” The claim, said NPR, has been “frequently refuted.” The Associated Press said the president was “wholly misrepresenting” the policy.

Even Trump-friendly Fox News noted the claim, then said “Here are the facts.” That alone prompted stories from The Hill and The Daily Beast.

Politicians’ repetition of false claims – even when fact-checkers have already called them on it – is both friend and foe of fact-checkers. It makes the checking faster and easier, since the work has already been done. But it can be disheartening because it suggests the politicians either don’t care that they’ve been debunked or aren’t paying attention.

The Washington Post’s Erik Wemple talked about this phenomenon in a column Tuesday about the repetition of falsehoods regarding the child separation policy. The Post last year  awarded Trump the “bottomless Pinocchio” award for repeating false claims, often about immigration.

What we liked: It may feel repetitive when a number of fact-checkers all focus on the same statement. But while fact-checkers are competitive, they often don’t worry about whether someone else has done a story. We think that’s a good thing.

In cases like this, where a politician’s comments are so blatantly false, there is strength in numbers. A broad consensus among fact-checkers on the veracity of a claim can really drive the conclusion home among readers across the country.

  1. It’s no surprise that lack of good information is bad for democracy. But three academics, writing in The Washington Post, explain why. Their identification of a “danger zone” of transparency is particularly relevant today.
  2. The Los Angeles Times dispelled some of the misconceptions about the U.S.-Mexico border.
  3. Of all the takes on Trump’s windmills-cause-cancer comments, one of the most thorough was from BuzzFeed News.
  4. Africa Check is seeking a deputy director, following Noko Makgato’s promotion to director last month.
  5. Platforms have taken a number of steps to block misinformation about vaccines. The Associated Press put them to the test and found they came up short.  
  6. This browser extension lets people check whether or not someone actually tweeted a key phrase.
  7. Duke University is starting them young. Senior Bill McCarthy recently won an annual award for local journalism for his work with the North Carolina Fact-Checking Project.
  8. Daniel spoke to a comedian who duped several American media outlets into covering a bogus health company selling $80 used tissues.
  9. Nieman Lab wrote about how news organizations can use metadata to combat misinformation.
  10. BuzzFeed News reported on a Facebook page that ran almost 5,000 ads for a bogus rebate scam.

That’s it for this week. Feel free to send feedback and suggestions to factchecknet@poynter.org.

Daniel and Susan