Frontier models today are either trained centrally and available behind paid API's, or trained centrally and opensourced. There appears to be the possibility of a third approach; Decentralized Learning, where models are sharded across nodes and exist within an elastic pool of independently controlled compute. This setting introduces significant technical challenges, however if instantiated such an approach would significantly alter the landscape of frontier model risk due to both novel the governance structures introduced and potentially unprecedented scale. To date, there has been no analysis of the risks such an approach introduces. We conduct this analysis and argue that the decentralized approach reduces rather than increases Frontier Risk. Additionally, we summarize the technical literature and conclude Decentralized Learning may be significantly more feasible than researchers are currently aware.