Algorithms surround people constantly in the digital world, yet their concrete functioning remains unknown to both researchers and users. Therefore, researchers have detected different mental models and folk theories through which laypeople try to explain algorithmic functioning. Yet, it is unclear what abstract concepts laypeople associate with algorithms that determine the online content that is displayed to them. Following a two-dimensional analysis of open-ended responses within a survey (N = 2,117), we clustered social media users’ mental associations as cognitive and behavioral and compared them between users from Germany and from the United States. While users in both countries had a general understanding of algorithmic automatic functioning, users from the United States expressed more cognitive associations. Across cultures, users do not perceive themselves as active in engaging with algorithms, with U.S. users using even more passive language. This study extends our knowledge about human understanding of algorithms, as it (a) highlights numerous cognitive associations, (b) analyzes the role of human–algorithm interaction, and (c) compares mental associations across cultures.