Corruption is widely defined as the abuse of public office for private gain. This definition has several terminological shortcomings and encourages anti-corruption reformers to design strategies in line with principal-agent solutions that usually target government officials and not others. Despite the widespread prevalence of this framework—and the massive transparency and accountability anti-corruption campaigns that have followed—corruption continues to afflict governments and societies globally. This is because corruption presents something more complex than a principal-agent problem. Unlike many other kinds of illegal or immoral acts, it carries steep collective political costs that have the ability to derail typical principal-agent reform efforts. Corruption should therefore be redefined as an event that occurs when an actor seeks an unauthorized benefit from an organization in a manner that could compromise the public’s trust in that organization. By focusing on the effects of the corrupt act on trust in organizations, this new definition encourages anti-corruption strategies that take account of broader considerations beyond simply punishing the ‘agent.’