This paper examines how the European Union’s Digital Services Act (DSA) understands platform addiction and, specifically, how it conceptualises its causative phenomenology: namely, the processes through which addiction is thought to arise. It begins by analysing the Act’s architecture and contends that the DSA operates with a largely mechanical, linear view: addiction is treated as the outcome of particular interface features that trigger or drive harmful behaviour. The paper then sets out four dominant models for understanding platform addiction and, drawing on Karen Barad’s concepts of intra-action and diffraction, argues that addiction is better approached as an emergent phenomenon produced through entangled human, social, and technical relations. On that basis, the paper’s central claim is not simply that the DSA simplifies; it is that the way it simplifies does not stand outside the phenomenon it regulates. By fixing predicates, thresholds, and evidential tests, the Act makes agential cuts that shape what counts as harm, when intervention is permitted, and how risk is distributed. These cuts are not neutral. They have distributional effects and could be drawn differently (through alternative predicates or thresholds) to produce other evidential routes and earlier intervention points, with different consequences for how risks and harms are allocated and managed.