WA

Mozilla is demanding that the world’s second-largest social media platform stem election disinformation through warning labels and other updates

(TUESDAY, APRIL 2, 2024) — Today, Mozilla is urging WhatsApp to reform key features to combat political disinformation that could threaten elections. It’s imperative that these changes be made in the next three months, ahead of major elections in India, South Africa, Mexico, the EU, and elsewhere.

With more than 2 billion users globally, WhatsApp is the world’s second-largest social media platform — but it frequently shirks the responsibilities that come with that. Mozilla’s research has revealed political actors routinely exploit WhatsApp’s features to distribute dangerous election disinformation and harass women and minority groups. Deepfaked video and audio, inaccurate political reporting by junk news sites, incendiary propaganda or hate speech, and conspiracy theories can spread across the app rapidly and unchecked. Meanwhile, WhatsApp makes significantly fewer election interventions than Facebook and Instagram.

Says Nicholas Piachaud, Director of Campaigns at Mozilla: “What WhatsApp claims to be, and what it actually is, are significantly different. And that gap may have major consequences for democracy. WhatsApp says it’s a closed messaging app, yet the platform has several public broadcasting features, like groups for over 1,000 people, stories, channels, and communities.”

Piachaud adds: “Historically, WhatsApp’s election interventions have been much too little, much too late. Without these immediate changes, hundreds of millions of voters over the next three months could be exposed to unprecedented levels of election disinformation.”

__________________________________________

Specifically, Mozilla is calling on WhatsApp to:

  • Add friction to forwarding messages: Reduce the ease with which messages can be forwarded on the platform by adding one additional step that nudges users to pause and reflect before they forward content.

  • Add disinformation warning labels to viral content: Automatically add clear “potential disinformation” warning labels to viral messages, in addition to the “forwarded many times” label currently in use.

  • Reduce WhatsApp’s broadcast capabilities: Disable the Communities feature and also limit the size of broadcast lists to 50 people and cap their usage to twice a day.

Mozilla is calling for these measures to be implemented for a fixed period of time: during polling days and in the month before and the month after the election.

__________________________________________

Unless WhatsApp changes key product features, political disinformation and other harmful content could threaten the integrity of upcoming elections around the world. That includes major elections happening over the next three months in India (starting April), South Korea (April), South Africa (May), Mexico (June), and the EU (June).

Global Majority countries are especially vulnerable to election disinformation spreading on messaging apps like WhatsApp, since these platforms are a core source of internet access.

This campaign is part of Mozilla’s broader 2024 election work, which entails research into synthetic content detection, the tracking of platforms’ election policies, and more.

Past WhatsApp failures

  • In the Brazilian 2022 general elections, WhatsApp served as a mass-broadcast tool for “participatory disinformation” in public groups. Audiences amplified junk news sources and co-created conspiracy theories with the goal of overturning the actual election results.

  • During the 2019 Indian general elections, political actors built lists of voters that included sensitive information like caste, religion, National ID and phone numbers, and created targeted WhatsApp groups to send these voters tailored political messages — as well as hate speech and disinformation.

  • Mozilla’s systematic review of WhatsApp’s elections announcements found that many announcements outside of the U.S. and EU appeared to be functionally the same boilerplate text — raising questions about the extent to which the platform is meaningfully adapting its approach to different political contexts.