Twitter is introducing a new strategy for addressing misinformation, one centered around community participation.

The Birdwatch initiative was initially discovered last autumn by tech analyst Jane Manchun Wong. Twitter has now initiated a trial phase of the project through the Birdwatch website.

According to a blog post from Keith Coleman, Twitter’s Vice President of Product, the intention is to move beyond the current system of labels applied to tweets that are disputed or potentially inaccurate. He indicated that these labels are currently reserved for “situations where content violates our policies or gains significant public attention.”

Coleman explained that the Birdwatch system is designed to “include a wider variety of viewpoints in addressing this challenge.” This expanded perspective will move beyond simply determining whether a tweet is factually correct, and could potentially lessen the pressure on Twitter regarding specific content moderation choices.

Individuals can register on the Birdwatch platform to identify tweets they believe contain misleading information, provide additional information through explanatory notes, and evaluate the usefulness of notes submitted by other users. Currently, these notes are only displayed on the Birdwatch website, but the company intends to integrate them into the primary Twitter platform.

“We are confident that this method can enable a swift response to the spread of inaccurate information, offering context that users find trustworthy and beneficial,” Coleman stated. “Our ultimate goal is to display these notes directly on Tweets for all Twitter users, once a broad and diverse group of contributors reaches a consensus.”

Considering the likelihood of debate and disagreement surrounding controversial tweets, it remains to be determined how Twitter will present these notes in a clear and manageable way, and how it will avoid becoming entangled in these discussions. The company has announced that Birdwatch will utilize algorithms based on “reputation and consensus systems” to rank content, and the code for these systems will be made publicly available. (All contributions to Birdwatch will also be accessible for download.) Further details regarding the initial ranking system can be found here.

“We recognize the numerous obstacles involved in creating a community-based system like this – from protecting it against manipulation to ensuring it isn’t controlled by a simple majority or skewed by the demographics of its contributors,” Coleman said. “We will be concentrating on these issues throughout the pilot program.”