This paper studies the cooperative output regulation problem for a class of nonlinear multi-agent systems with unknown control directions subject to jointly connected switching networks. Our approach integrates three techniques, namely, the distributed internal model, the Nussbaum gain technique, and the adaptive distributed observer. As a result, an adaptive distributed control law is synthesized based on the certainty equivalence principle. In contrast with the existing results, our approach leads to the solution of the cooperative output regulation problem for a class of nonlinear multi-agent systems whose control directions are unknown and can be different and whose communication network can be directed and disconnected at every time instant.