Birdwatch: Twitter rolls out system for people to annotate incorrect tweets

Birdwatch is ‘not a place for quick dunks, personal opinions, or insults,’ Twitter says

(AFP via Getty Images)

Twitter has announced Birdwatch, a program whereby users can identify tweets that contain misinformation and provide context in an attempt to stop their spread.

It comes as the company faces continued criticism for the spread of false information, resulting in the company adding warning labels to tweets and removing former president Donald Trump from the platform– a key figure for spreading misinformation.

The pilot program unveiled Monday allows a preselected group of users — for now, only in the US — who sign up through Twitter. Those who want to sign up must have a US-based phone carrier, verified email and phone number, and no recent Twitter rule violations.

"It's not a place for quick dunks, personal opinions, or insults," the site explains.

"Write clear, evidence-based notes to help everyone, even sceptics, better understand a tweet and why it might be misleading."

Twitter said it wants both experts and non-experts to write Birdwatch notes. It cited Wikipedia as a site that thrives with non-expert contributions.

"In concept testing, we've seen non-experts write concise, helpful and easy-to-understand notes, often citing valuable expert sources," the company wrote in a blog post.

Twitter, along with other social media companies, has been grappling how best to combat misinformation on its service. Despite tightened rules and enforcement, falsehoods about the U.S. presidential election and the coronavirus continue to spread.

But if the effort is to work, Twitter will have to anticipate misuse and bad actors trying to game the system to their advantage.

To help weed out unhelpful or troll-created notes, for instance, Twitter plans to attach a "helpfulness score" to each one and will label helpful ones "currently rated helpful."

The company said Birdwatch will not replace other labels and fact checks Twitter currently uses — primarily for election and COVID-19-related misinformation and misleading posts.

The program will start with 1,000 users and eventually expand beyond the US. Should the trial be a success, the social network aims to make notes appear directly on tweets for the global Twitter audience.

San Francisco-based Twitter said it is trying to ensure that Birdwatch has a diverse range of perspectives and participants — an ongoing problem at Wikipedia, where many of the contributors and editors are white men.

"If we have more applicants than pilot slots, we will randomly admit accounts, prioritizing accounts that tend to follow and engage with different audiences and content than those of existing participants," Twitter wrote.

Additional reporting by agencies