A little over a week after the U.S. election, Twitter has revealed some numbers, including that it has labeled over 300,000 tweets as election misinformation so far.
“Approximately 300,000 tweets have been labeled under our Civic Integrity Policy for content that was disputed and potentially misleading,” wrote policy lead Vijaya Gadde and Periscope co-founder Kayvon Beykpour in a blog post.
“These represent 0.2% of all U.S. election-related tweets sent during this time period.”
The company also disclosed 456 of those flagged tweets were also covered with a warning message in addition to limiting their engagement features. Those likely include a series of tweets from the incumbent U.S. President Donald Trump, even though Twitter did not detail any specific examples.
As a result of these measures, the company adds that 74% of people saw these tweets only after Twitter had already labeled them for misinformation.
It also reported there was a 29% increase in quotes for the tweets in question.
Twitter’s eagerness to moderate election discussions also had some negative outcomes.
On the blog post. the company said that its decision to disable “liked by” and “followed by” recommendations from accounts that users don’t follow “prevented many people from discovering new conversations and accounts to follow.”
"While we had initially hoped that this would help reduce the potential for misleading information to spread on our service, we did not observe a statistically significant difference in misinformation prevalence as a result of this change," Twitter said.
For this reason, the micro-blogging service planned to be bring the feature back.
It should also be noted that Twitter, and a number of other social media platforms, can editorialize contents published by third-parties on its platform because they are granted the protections engendered by the contentious U.S. legislation Section 230 of the Communications Decency Act, which protects U.S. companies from being sued for content on their platform.
