In October 2020, Twitter launched Birdwatch, a crowdsourced fact-checking program that tapped a group of about 10,000 volunteers. The beta program had the volunteers write notes to provide context on existing tweets that had the potential to be misleading. They were also asked to rate those notes to improve the performance of their annotation efforts.
Now, Twitter has announced that the notes created by members of its Birdwatch pilot program will be visible to additional users across the US. The notes will appear "directly on some tweets," where these new participants will be able to rate those notes and provide feedback on their efficacy to the Birdwatch program. The actual group writing these notes will not be expanded, however.
For a note to appear to users within the expanded program, it must first be rated as "helpful" by "enough Birdwatch contributors from different perspectives." Twitter said it is building out perspective profiles for all of the program's participants by compiling how they have rated notes in the past. It claims it is attempting to cultivate a diverse range of perspectives across topics, and it promises that it does not take any user's demographics into account.
More: YouTube has removed over 1 million coronavirus misinformation videos
Obviously, Twitter's decision to expand the program comes as a result of a successful early pilot. The company detailed some aspects of that success in its new announcement, revealing that the "majority of people found Birdwatch notes (specifically those designated helpful by Birdwatch contributors) helpful," including users from "across the political spectrum."
The overall effect was a 20% to 40% drop in the likelihood that a users would "agree with the substance of a potentially misleading Tweet after reading a note about it." This is compared to users reading the same tweet without the additional context of Birdwatch annotation.
Twitter hopes Birdwatch can be an impactful tool in its constant war against its platform serving as a dispenser of misinformation and malicious propaganda. This expansion of the pilot program will loop in a "small (and randomized) group of people on Twitter in the US, meaning the program has a long way to go before it can positively impact Twitter's entire US userbase, let alone its global audience.