Twitter said Thursday that it labeled roughly 300,000 posts as containing content that was “disputed and potentially misleading” during a two-week period around the general election.
Those tweets accounted for roughly 0.2 percent of all tweets related to the election in the period spanning Oct. 27 to Nov. 11.
Four-hundred and fifty-six of those 300,000 tweets had interstitial labels placed on them that required users to click through to read the posts.
Roughly 74 percent of users who saw the flagged tweets viewed them after the label was applied, Twitter said.
"These enforcement actions remain part of our continued strategy to add context and limit the spread of misleading information about election processes around the world on Twitter," Vijaya Gadde and Kayvon Beykpour, who respectively lead Twitter's legal and product teams, wrote in a blog post on Thursday.
President TrumpDonald TrumpFormer Sen. Heller to run for Nevada governor Overnight Defense & National Security — Milley becomes lightning rod Joint Chiefs Chairman Milley becomes lightning rod on right MORE’s account was one of the ones hardest hit by Twitter’s labeling.
Between Election Day and Nov. 7 — the day former Vice President Joe BidenJoe BidenOvernight Defense & National Security — Milley becomes lightning rod Democrats hope Biden can flip Manchin and Sinema On The Money — Presented by Wells Fargo — Democrats advance tax plan through hurdles MORE was projected to win the race — nearly half of Trump’s election-related tweets had been labeled.
Twitter also announced Thursday that it is keeping some of the changes it made for the election in place for the foreseeable future.
Users will continue to be prompted to add context to tweets instead of just retweeting them moving forward.
Twitter said that that feature reduced the sharing via both quote tweets and retweets by 20 percent and helped reduce the spread of misinformation.
"This change introduced some friction, and gave people an extra moment to consider why and what they were adding to the conversation," the company officials wrote in their blog post. "In short, this change slowed the spread of misleading information by virtue of an overall reduction in the amount of sharing on the service."
The platform will undo a couple of changes, resuming recommending tweets from people that users don’t follow and ending the requirement that trends must have explanations before they can land in “For You” suggestions.