Many studies have shown that stimuli modulate neuronal responses, but responses are also affected by the state of the local network. However, little is known about how intrinsic fluctuations of population activity modulate the sensory tuning of cells and whether they affect their encoded information. We found that fluctuations of population activity in V1 modulate the tuning of individual neurons in a multiplicative and additive manner. Neurons with strong multiplicative effects tended to have little additive modulation, and the reversed was observed for neurons with strong additive effects. As predicted by a model based on a multi-gain model of population activity and using state-of-art decoding techniques from simultaneously recorded neuronal populations, we found that the information encoded by multiplicatively-modulated neurons increased with greater population activity, while that of additively-modulated neurons decreased. These effects cancel each other in such a way that fluctuations of population activity had little effect on total information. Therefore, our results suggest that the global state of the network acts as a 'traffic light' that controls which subset of neurons are most informative without affecting total information.