Twitter pilots 'potentially messy' crowd-sourced fact checking
Twitter has announced a new project that will recruit users of the social media platform to help tackle false and misleading tweets. The company said: "We know this might be messy... but we believe th
Twitter has announced a new project that will recruit users of the social media platform to help tackle false and misleading tweets.
The company said: "We know this might be messy... but we believe this is a model worth trying."
The crowd-sourced pilot programme, named Birdwatch, is being tested in America and follows criticism of social media companies in the wake of the mob attack on the US Capitol building.
The "community-driven approach" will initially allow a select group of users to add notes providing additional context to misleading tweets, with the aim that these notes will contribute to building understanding.
"It's not a place for quick dunks," the company warned.
At first these notes will be visible on a separate Birdwatch site, but in time the company plans to make the notes directly visible on Twitter for its global audience.
Twitter said it has conducted more than 100 interviews with "individuals across the political spectrum" who have provided "broad general support for Birdwatch".
"In particular, people valued notes being in the community's voice (rather than that of Twitter or a central authority) and appreciated that notes provided useful context to help them better understand and evaluate a Tweet (rather than focusing on labelling content as 'true' or 'false').
"Our goal is to build Birdwatch in the open, and have it shaped by the Twitter community," the company said.
Birdwatch will be powered by computing a "smoothed helpfulness score" for each note, which is Twitter says "simply the proportion of ratings that say the note is helpful, smoothed by adding five to the denominator".
"Then out of all notes that have received at least three ratings, the top three by 'smoothed helpfulness score' are annotated as 'currently rated helpful' as long as the 'smoothed helpfulness score' is at least 0.5."
The move follows controversy over the social media platform's response to misinformation in the wake of Donald Trump's presidency and the mob attack on the Capitol.
Although it suspended 70,000 accounts linked to the far-right conspiracy movement QAnon in light of the violence, Twitter faces additional scrutiny during Donald Trump's second impeachment trial.
The social media company and the Trump administration repeatedly clashed during the president's final months in office, prompting the White House to drive forward changes to social media regulations.
The company was accused of censorship by those sympathetic to Mr Trump and of failing to address his inciting messages by those critical of him.
Just days before the inauguration of Joe Biden, Twitter confirmed that it was permanently banning Mr Trump from its platform.