Algorithmic advertising is the subject of a huge range of conflicting narratives. On the one hand, it seems to be the embodiment of surveillance capitalism: encouraging mass data collection, facilitating discriminatory treatment, and affecting how we form preferences and express ourselves online. On the other hand, it also allows for the personalized matching of consumer preferences with relevant products and services, providing the financing for our modern, innovative, mostly free digital economy. These harms and benefits are not equally distributed across populations. This Note argues that the way in which these competing narratives overlap and coexist reveals that algorithmic advertising is a \u201cwicked problem\u201d given
\nthe pervasive opacity of the AdTech ecosystem, the presence of multiple stake-holders and decision-makers with conflicting values, and systemic complexity.
\nThis Note does not aim to take a clear position on the normative desirability of targeted advertising or to propose a straightforward solution that easily resolves these complexities. Instead, I argue for a form of regulation that I term \u201ctransparency as infrastructure regulation.\u201d Before stakeholders can form any substantive policy responses, there must be a baseline level of transparency around algorithmic advertising to create an empirical knowledge base for societal stakeholders to collectively create policy interventions, which I refer to as a \u201ccollective perspective\u201d on targeted advertising.<\/p>\n