Any message (text, image, video) delivered on a social media platform that reminds users of the importance of considering the accuracy of news.

By calling their attention to accuracy, users will be more apt to discern if a news item is misinformation and less likely to re-share fake news.

Several survey experiments (Pennycook & Rand, 2022), including one large multinational study (Arechal et a., 2023), demonstrate that accuracy nudges could be an effective way to reduce the spread of misinformation online; they show that if you first prompt people "in the lab" to think of accuracy, they will then be less likely to say they would share articles that are fake news.   

Capraro and Celadin (2022) operationalize how accuracy nudges might be applied online, likewise conducting several online lab experiments.  Participants were shown true and false news items and given the option to click on a "share" button to indicate they would share them on social media. When the researchers included the text "think if this news is accurate" on the share button, participants were more than 25% less likely to indicate they'd share fake news compared to participants who did not see an accuracy prompt. (Note: all effects we include are statistically significant, unless otherwise stated.) Importantly, the accuracy reminder did not decrease participants' intention to share real news.

Most promisingly, Lin et al. (2024) tested accuracy nudges in the field, using ads on Facebook and Twitter/X to deliver a variety of text, image and video ads to a set of users ("treatment) and observing how much less likely they were to share misinformation compared to a set of other ("control") users. Among users who recently shared misinformation in the treatment group, the researchers observed ~6% decrease in shares of misinformation, compared to the control group. They observed no difference among users who had not recently shared misinformation, however. As in Capraro and Celadin (2022), the authors likewise did not see a drop in how much users shared other (non-misinformation) news. The authors also observed no difference in the type of accuracy nudge used, which included "tips" for detecting fake news, a poll asking users about the importance of accuracy, or a message that most people value accuracy. While the results of this paper are strongly encouraging, it is important to note that given the way ads are distributed on platforms (i.e. targeting users more likely to be receptive to them) and that the authors were unable to know who observed the ads, it is possible that accuracy nudges would be ineffective - or even have a backfiring effect - among a subset of users. 

Fake news makes up a small percentage of information on social media, but because fabricated stories are often designed to provoke division and anger, even small amounts of fake news can lead to civic and physical harm. Platforms may be limited in their ability to identify and remove fake news, however, because of the challenges in creating reliable algorithms and concerns over censorship. By prompting accuracy, platforms can call on the judgment of good faith users to identify and slow the spread of fake news.

In another version of their intervention, Capraro and Celadin include the text "I think this news is accurate." In this "accuracy endorsement" prompt, participants were even less likely to share fake news and significantly more likely to share real news. While an "endorsement" of news accuracy goes farther than merely prompting users to think of accuracy and changes the experience of sharing, it may be an effective way to encourage users to promote accurate news. It also may increase engagement because, as mentioned above, real news makes up the vast majority of links shared on social media.