Election misinformation a problem in any language

WASHINGTON — Warnings about deepfakes and disinformation fueled by artificial intelligence. Concerns about campaigns and candidates using social media to spread lies about elections. Fears that tech companies will fail to address these issues as their platforms are used to undermine democracy ahead of pivotal elections.

Those are the worries facing elections in the U.S., where most voters speak English. But for languages like Spanish, or in dozens of nations where English isn’t the dominant language, there are even fewer safeguards in place to protect voters and democracy against the corrosive effects of election misinformation. It’s a problem getting renewed attention in an election year in which more people than ever will go to the polls.

Tech companies have faced intense political pressure in countries like the U.S. and places like the European Union to show they’re serious about tackling the baseless claims, hate speech and authoritarian propaganda that pollutes their sites. But critics say they’ve been less responsive to similar concerns from smaller countries or from voters who speak other languages, reflecting a longtime bias toward English, the U.S. and other western democracies.

Recent changes at tech firms — content moderator layoffs and decisions to rollback some misinformation policies — have only compounded the situation, even as new technologies like artificial intelligence make it easier than ever to craft lifelike audio and video that can fool voters.

These gaps have opened up opportunities for candidates, political parties or foreign adversaries looking to create electoral chaos by targeting non-english speakers — whether they are Latinos in the U.S., or one of the millions of voters in India, for instance, who speak a non-english language.

“If there’s a significant population that speaks another language, you can bet there’s going to be disinformation targeting them,” said Randy Abreu, an attorney at the U.s.-based National Hispanic Media Council, which created the Spanish Language Disinformation Coalition to track and identify disinformation targeting Latino voters in the U.S. “The power of artificial intelligence is now making this an even more frightening reality.”

Many of the big tech companies regularly tout their efforts to safeguard elections, and not just in the U.S. and E.U. This month Meta is launching a service on Whatsapp that will allow users to flag possible AI deepfakes for action by fact-checkers. The service will work in four languages — English, Hindi, Tamil and Telugu.

Meta says it has teams monitoring for misinformation in dozens of languages, and the company has announced other election-year policies for AI that will apply globally, including required labels for deepfakes as well as labels for political ads created using AI. But those rules have not taken effect and the company hasn’t said when they will begin enforcement.

The laws governing social media platforms vary by nation, and critics of tech companies say they have been faster to address concerns about misinformation in the U.S. and the E.U., which has recently enacted new lawsdesigned to address the problem. Other nations all-too often get a “cookie cutter” response from tech companies that falls short, according to an analysis published this month by the Mozilla Foundation.

The study looked at 200 different policy announcements from Meta, Tiktok, X and Google (the owner of Youtube) and found that nearly twothirds were focused on the U.S. or E.U. Actions in those jurisdictions were also more likely to involve meaningful investments of staff and resources, the foundation found, while new policies in other nations were more likely to rely on partnerships with fact-checking organizations and media literacy campaigns.

Odanga Madung, a Nairobi, Kenya-based researcher who conducted Mozilla’s study, said it became clear that the platforms’ focus on the U.S. and E.U. comes at the expense of the rest of the world.

“It’s a glaring travesty that platforms blatantly favor the U.S. and Europe with excessive policy coddling and protections, while systematically neglecting” other regions, Madung said.