Twitter Expands Crowdsourced Fact-Checking Following Phony Viral Photo

About a week after an artificial intelligence (AI) generated image of an explosion outside the Pentagon went viral on Twitter, the social media giant is using its crowdsourced fact-checking program, Community Notes, to fight back against fake or misleading images.

“Community Notes aims to create a better informed world by empowering people on Twitter to collaboratively add context to potentially misleading Tweets,” Twitter writes.

Contributors can leave notes on any Tweet, and if people rate those notes as helpful, they can even become publicly shown as attached to a popular Tweet. For example, Community Notes can add context to posts, and even label Tweets as false.

However, these crowdsourced notes, which act as a check on people writing misleading and false information, have been limited to text. With today’s announcement, contributors can now add information specifically related to an image. A new option allows contributors to make notes about pictures, and these notes can then appear in all Tweets that include the same image.

As the viral fake Pentagon incident showed last week, AI-generated images have reached the point now where people can’t tell fact from fiction. In some cases, AI-generated images cause little harm, but in the Pentagon case, the phony photo had a real-world impact on the stock market. The image created a $500 billion market cap swing. Although the economic impact was short-lived, it was no less real — and all from a fake photo on social media.

Currently, Community Notes are limited to single images, and can’t be applied to videos and posts, including multiple photos. However, per The Verge, Twitter is working on that functionality, which will undoubtedly be important as AI-based video generation, while somewhat lackluster now, will surely improve.

Twitter’s owner and CEO, Elon Musk, recently joined with other tech experts to call for a pause on developing advanced AI technology.

Community Notes, previously called Birdwatch, have proven helpful in preventing users from retweeting false information, and will presumably have a similar effect concerning Tweets with images. According to Twitter, users are 15 to 35 percent less likely to like or retweet a Tweet that has an attached note compared to a person who sees the Tweet without its crowdsourced note.

Twitter’s corporate partners no doubt took note of last week’s Pentagon debacle and the accompanying stock market effects. Other fake photos have gone viral on Twitter, albeit without economic impact. Cynicism concerning the impetus behind the new policy aside, any action aimed at curtailing the spread of misinformation is a step in the right direction.

Image credits: Twitter