My dissertation explores prejudice, particularly what makes prejudices resistant to change, and the impact of prejudice in the world, such as how prejudice influences our norms and practices of privacy. In the first chapter, I engage with a common view of prejudice, ‘the Bad Emotion view,’ which claims prejudices are negative stereotypes driven by bad emotions. These bad emotions cause the agent to act irrationally in the face of counterevidence and partly explain prejudice’s epistemic resilience. I argue that some prejudices don’t involve bad emotions at all, and many prejudices involve emotional complexity. A common assumption about prejudices is that they involve an epistemic mistake – namely, an unjustified belief. However, some who endorse the view that prejudices are stereotypes argue that prejudices can be justified. In my second chapter, I argue that the epistemic problem of prejudice is not a failure to justify beliefs (though many prejudices will be unjustified). To find the epistemic mistake involved in prejudice, we need to look beyond stereotypes and their logical structures: we need to look to the larger belief structure that the stereotype is housed in, and some prejudices might not even involve stereotypes at all. Instead, I argue that the epistemic problem with prejudices is their epistemic resilience. I advance a novel account of prejudice which holds that they are conceptual frameworks — networks of beliefs, methods, assumptions and concepts. I argue that these conceptual frameworks of prejudice are motivated by supporting an agent’s (perceived) interests. In the third chapter, I look into the connection between privacy and social epistemology. I argue that a prominent theory of privacy developed by Helen Nissenbaum, Contextual Integrity, does not account for how unjust social power relations shape the norms and practices surrounding privacy. I argue that privacy violations are both more pervasive and harmful for marginalized groups. Many privacy violations are caused by epistemic injustices, such as, when distrust for a marginalized group causes someone to look for more information. I claim there is a particular kind of privacy violation caused by epistemic injustice which I call ‘knowledge entitlements.’