As of November 2022, Birdwatch is now Community Notes.
We’ve been piloting our Birdwatch program for over a year to bring helpful context to Tweets in a transparent and collaborative way.
Now for the first time, we’re expanding the visibility of Birdwatch notes — written and selected by people on Twitter, for people on Twitter — to everyone in the US.
Starting today, if you’re using Twitter in the US, you’ll begin to see some Tweets accompanied by a note containing relevant information that’s been rated “Helpful” by Birdwatch contributors. Most notes contain additional sources that can be clicked for an even deeper dive into a subject or conversation.
You'll also have the ability to rate the notes you see to help us understand if they’re helpful or not.
Birdwatch is made up of independent contributors, and individual notes are never written by Twitter, Inc. This is intentional, as it helps ensure our efforts to address potentially misleading information are informed by a diverse group of people who use our service.
The program is designed to surface notes that are informative and helpful to as many people as possible thanks in large part to what’s known as a bridging algorithm. Here’s how it works:
In order to be shown on a Tweet, Birdwatch notes need to be found helpful by people who have tended to disagree in their past ratings. This means the algorithm takes into account not only how many contributors rated a note as Helpful or Not Helpful, but also whether people who rated it seem to come from different perspectives. Check out our Birdwatch guide for more info on this.
In many online spaces, especially those using engagement-based ranking, divisive content can be more likely to go viral. Bridging-based ranking systems aim to overcome this 'bias toward division'. Birdwatch’s use of bridging to elevate context found helpful by people that tend to disagree is an exciting step toward a better internet — one that supports those building common ground.
Additionally, to keep the quality of notes high and encourage contributors to write and rate notes in a thoughtful way, we developed a Birdwatch onboarding process through which new contributors earn the ability to write notes by consistently rating other contributors’ notes and reliably identifying those that are Helpful and Not Helpful.
We’re opting to expand the visibility of Birdwatch notes to everyone in the US based on research that indicates the program is helpful and informative to people who hold different points of view.
According to the results of four surveys run at different times between August, 2021 and August, 2022, a person who sees a Birdwatch note is, on average, 20-40% less likely to agree with the substance of a potentially misleading Tweet than someone who sees the Tweet alone. Survey participation ranged from 3,000 to more than 19,000 participants, and the results were consistent throughout the course of the year, even as news and Tweet topics changed.
We also see that Birdwatch informs sharing behavior. Analyzing our internal data, we’ve found that a person on Twitter who sees a note is, on average, 15-35% less likely to choose to Like or Retweet a Tweet than someone who sees the Tweet alone.
In our most recent survey, notes were found to be informative regardless of a person's self-identified political party affiliation — there was no statistically significant difference in average informativeness across party ID.
We believe that earning trust starts with transparency. That’s why we’ve made the Birdwatch algorithm publicly available on GitHub, along with all the data that powers it, so anyone can audit, analyze, or suggest improvements. Whether you’re a researcher, a hobbyist, or simply curious, you can explore and run the Birdwatch algorithm on your own computer using the guide link shared above.
You can also keep up to date with all things Birdwatch by following our official @Birdwatch Twitter account, or click this link to apply to become a Birdwatch contributor and help improve the health of conversations on Twitter.