All social media users have interacted with an algorithm, but there’s not much transparency about how these algorithms decide what shows up on people’s feeds. While users have a role in shaping part of their social media feeds, it’s often advertisers, politicians, and news media organizations that shape the content they’re exposed to in ways users don’t always realize. Social media algorithms, like Facebook’s, tend to categorize users into groups based on their preferences, but these generalized models can create mistakes at the individual level. Those who are uninterested in political content, as classified by the Facebook algorithm, are often left out of political discourse online, and it’s difficult for them to change these preferences because they’ve already been designated into a strict category (Thorson et al, 2019). Dr. Khan said algorithms on social media platforms must narrow down large amounts of content to decide what shows up on users’ feeds which can create issues. “Because of this lack of transparency by platforms, some news, which could be local news and could be very important to us, may not ever show up in my feed because I’m clicking on other things” (L. Khan, November 17, 2025). Khan calls these “media deserts” that stem from the decisions made by algorithms that lead to the erasure of important content. Additionally, politicians and partisan organizations use these categories to target specific groups on Facebook to sway voter behavior. Facebook’s algorithms decide what content shows up on a user’s feed based on the predicted relevance to the user which is part of the app’s business model and allows it to sell data to advertisers. These advertisers then use this user data to target specific audiences, sometimes politically motivated, to attempt to influence people through their social media feeds. Facebook users’ online behavior is used to infer information about their political interests, but the algorithm it curates for the user is not always accurate or unbiased (Thorson et al, 2019).
X has a timeline that is completely selected and ordered by algorithms. The “Home” timeline is organized by a personalized relevance model, similar to Facebook, and amplifies content that it feels is most relevant to the user. However, a study done at Stanford University found that the algorithms on X tend to promote more right-wing content than left-wing content. Their research found that conservative politicians benefit more from the X algorithm than other politicians (Huszár et al, 2021). This could mean that X users in the United States are more likely to see Republican content on their feeds than any other type of political content. This study points to the theory that X could be Agenda Setting on behalf of the Republican party, and promoting content that it wants to be on the forefront of people’s minds. The X algorithm could be biased toward conservative ideology and may be trying to push a certain agenda onto its users. The study also found that the X algorithms are more likely to amplify sources, such as media content, that are more partisan compared to media outlets that are more moderate (Huszár et al, 2021). The authors were concerned that more biased news reporting is more likely to have higher amplification on X, leading users to be exposed to more extreme ideologies rather than mainstream political voices. Dr. Khan’s research also led him to reach the same conclusion. He found that algorithms are optimized for engagement, so more dramatic content can receive higher amplification than accurate media sources. “Whatever is going to be sensational, polarizing, and emotionally charged is going to show in our feeds” (L. Khan, November 17, 2025).
The YouTube algorithms has also been found to amplify and promote right-wing ideologies. While this radical conservative content is already more likely to attract people who already have a level of racial bias, these people are also more likely to be recommended extremist videos by the YouTube algorithm (Epstein, 2024). YouTube users who are already politically right-leaning can more easily be indoctrinated in extremist right-wing content and can easily fall into YouTube rabbit holes by following suggested videos from the platform. Algorithms make it easier for people to get caught in “echo chambers” that reinforce and polarize their existing political opinions. Dr. Khan said people tend to feel strongly about the group or community that they belong to which leads them to believe in false content or promote it due to loyalty. “Users share misinformation not because they believe in it, but because it reinforces their group identity, sense of belonging, and their emotional connection to that content” (L. Khan, November 17, 2025). The YouTube platform also has an advantage over Facebook and X because it’s all video-based. Videos are more influential than text in convincing people that political content is truthful, and videos tend to get far more shares than other forms of media. The experiment Epstein and his team conducted also showed that YouTube, as a platform, can have a significant impact on voters through the strategic ordering of videos. He found that the videos suggested by YouTube through its algorithm can dramatically shift the opinions and voting preferences of undecided voters (Epstein, 2024). Whether users realize it or not, the YouTube algorithm is using a form of symbolic interactionism through its recommended videos. If users see more videos with conservative ideology, they’re going to view the world with those biases because that’s the only thing they’re exposed to. For example, if a YouTube user is only consuming content with racist undertones, they’re more likely to stereotype a black man as dangerous because that’s what they’ve been conditioned to think. Epstein found that people are also typically unaware that they are being politically influenced because this manipulation is disguised as recommendations based on their own video preferences (Epstein, 2024). Social media sites, whether we’ve noticed or not, have become corrupted by outside influences, specifically large corporations hoping to unconsciously influence our beliefs.