A perspective on friction interventions to curb the spread of misinformation.

Social media has enabled the spread of information at unprecedented speeds and scales, and with it the proliferation of high-engagement, low-quality content. Friction-behavioral design measures that make the sharing of content more cumbersome-might be a way to raise the quality of what is spread online. In this perspective, we propose a scalable field experiment to study the effects of friction with a learning component to educate users on the platform's community standards. Preliminary simulations from an agent-based model suggest that while friction alone may decrease the number of posts without improving their quality, it could significantly increase the average quality of posts when combined with learning. The model also suggests that too much friction could be counterproductive. Experimental interventions inspired by these findings would be minimally invasive.