Twitter announces plans to 'protect' users from 'harmful misleading information' ahead of midterms

Today, Twitter announced that it would take steps to "protect civic conversation" on its platform ahead of the midterm elections. Its plan includes cracking down on "harmful misleading information" by labeling, deleting, and turning off the ability for tweets to be replied to, liked, and shared.

Twitter stated that it is preparing to enforce its Civic Integrity Policy in the United States. It reported that this policy had been implemented during other global elections this year, including in the Philippines, Kenya, Australia, Brazil, and India.

The company said that these new guidelines will ensure online discourse does not "manipulate or disrupt civic processes."

As part of the policy, it will look for tweets that undermine confidence in an election or provide "misleading information" about how to get involved in an election. It will also crack down on suppression, intimidation, and fake accounts.

Twitter warned that tweets violating these rules might be deleted or receive a warning label. The company stated that it would reduce the visibility of or turn off the ability to like, share, and reply to tweets that violate its rules.

The Civic Integrity Policy stated, "In some instances, we'll also turn off your ability to reply, Retweet, or like the Tweet. We prioritize producing Twitter Moments in cases where misleading content on Twitter is gaining significant attention and has caused public confusion on our service. Labels applied to Tweets accrue 1 strike."

Twitter enforces a strike policy when deciding to either temporarily or permanently ban a user from the platform. Two and three strikes lock the user out of the account for 12 hours, four strikes lock a user out for a week, and five strikes will result in permanent suspension.

The social media platform reported the success of labeling tweets in the past. Twitter wrote, "Late last year, we tested new misleading information labels and saw promising results. The new labels increased click through rates by 17%, meaning more people were clicking labels to read debunking content. We also saw notable decreases in engagement with Tweets labeled with the new design: -13% in replies, -10% in Retweets and -15% in likes."

Twitter also assured users that it would bring back pre-bunks "to get ahead of misleading narratives on Twitter, and to proactively address topics that may be the subject of misinformation. Over the coming months, we'll place prompts directly on people's timelines in the US and in Search when people type related terms, phrases, or hashtags."