Twitter is bringing back 'prebunks' to get ahead of midterm misinformation

Twitter has announced a series of changes it hopes will "protect civic conversation" on the platform ahead of the U.S. midterm elections. Here's a rundown of how they'll do it.

First, they'll flag misinformation with labels that link to credible information or context. A test of labels last year decreased engagement with misinformation by 13 percent in replies, 10 percent in retweets, and 15 percent in likes. The labels also saw a 17 percent increase in clickthrough rates, a promising indication that more people were clicking to read "debunking content." The platform will also neither recommend nor amplify tweets it tags as misinformation and, in some cases, those tweets will not be able to be liked or shared.

Twitter will also be launching several product features, including prebunks, which are messages that appear at the top of users’ feeds to debunk false voting information. Prebunks were first used in 2020.

In addition to updating state-specific event hubs on the platform, a dedicated Explore tab will feature national news in English and Spanish by reputable news outlets (curated by Twitter), localized news and resources by state, and voter education public service announcements. 

And, last but not least, Twitter is ramping up security for public figures involved in elections. Candidate account labels will help users identify designated U.S. Senate, U.S. House of Representatives, or Governor race candidate accounts and tweets. And to protect and reinforce the security of accounts of government officials, candidates for office, journalists, and other groups, they're adding "sophisticated detections and alerts" around suspicious activity, login defenses, and expedited account recovery support.

For more information on Twitter's approach to elections, you can visit elections.twitter.com.