Technology governance does not happen in a vacuum. Platforms govern their users through design choices, content policies, and social norms, and they are governed by states, who seek to regulate the behavior of internet users via the moderation systems of platforms. On both sides of this relationship, institutions form to mediate governance practice: professional associations, auditing firms, moderation outsourcing companies, compliance consultants, and more. In this dissertation, I present three studies of these governance intermediaries and the markets that form around them. Rather than treating these markets as reactive, fully rational service providers who respond to a well-understood need from platforms and government, I treat them as active participants in constructing the markets they fill. In doing so, they reinterpret and alter the practical meaning of the law, shape the routines and norms of online safety professionals, and articulate what counts as a threat and an appropriate solution. Across my three studies, I examine how the emerging, VC-backed market for trust and safety software platforms mediates new AI and online safety laws; how militaristic, security-based language that describes disinformation as a technical, quantifiable threat opens a market for cybersecurity-esque solutions to social and political problems; and how third-party firms that automate copyright enforcement transform the practice of the notice-and-takedown regime of the Digital Millenium Copyright Act.