The Digital Services Act (DSA) is a critical piece of the procedural puzzle of generating a safe social media environment. This article highlights how the DSA’s ambitious provisions regarding content moderation rely on psychological assumptions and inferences about behaviour, yet do so in the absence of extensive empirical behavioural evidence. The difference between how individuals are assumed to behave and evidence on how they actually behave is a crucial element in the complex landscape of effectively moderating content on social media. For this reason, the article sheds light on the value of behavioural research in understanding preferences, incentives and decisions, and the role of such information in interpreting and developing legal provisions. Specifically, through explaining the results of a novel experimental study designed with the DSA in mind, the article offers insights on certain decision-making processes, and how this evidence can steer away from unfounded suppositions to help interpret and apply the DSA against the reality of behaviour.
