This research aims to develop a novel, philosophically grounded approach to the problem of misinformation on social media. Up to now, academics and companies have explored technical solutions such as debunking or suspending misinformation-spreading accounts. Similarly, the current body of research, both in philosophy and in the social sciences, is focused on the features inherent to misinformation itself. This leaves a significant gap in understanding the role of people who tend to engage with false claims. This research aims to address this gap by using natural language processing (NLP) techniques to examine the epistemic characteristics of misinformation consumers. The main thesis is that in order to combat misinformation on social media, we shouldn’t focus on the endless struggle of curbing false content but rather on the underlying epistemic vices (that is, harmful thinking patterns like closed-mindedness) that allow it to flourish. The project proposes a novel approach to combating misinformation by targeting the algorithmic design of social media platforms to promote the dissemination of accurate information and foster epistemic virtues (that is, beneficial thinking patterns) among users. The project has both a practical and a philosophical side. The practical side focuses on developing solutions that can be applied to combat misinformation on social media. The philosophical side of the project delves into the root causes of why misinformation spreads. It examines how social media platforms can encourage negative thinking patterns (epistemic vices) that make users more susceptible to believing false information. The project also considers the ethical implications of regulating social media platforms and argues that some restrictions on free speech can be justified if they improve the overall quality of public discourse.