Recent studies of media and security continue to be limited by various theoretical and ontological commitments, despite their claims of "moving beyond" or "reconsidering" the CNN effect. In this article, I review the two dominant research paradigms for media and security. The first research paradigm, which includes the CNN-effect literature, imagines media as an independent actor in the policy-making process. The second research paradigm portrays media as a neutral channel, passing along the message of foreign policy elites. Each of these paradigms remains wedded to an actor-centered and choice-theoretic approach to media and security, as witnessed by recent attempts to study the CNN effect. I argue that a new research paradigm based on relational sociology actually moves media and security studies beyond the CNN effect and resolves the fundamental theoretical limitations associated with it.