Adaptive behavior is based on flexibly managing and integrating perceptual and motor processes, and the reconfiguration thereof. Such adaptive behavior is also relevant during inhibitory control. Although research has demonstrated local activity modulations in theta and alpha frequency bands during behavioral adaptation, the communication of brain regions is insufficiently studied. Examining directed connectivity between brain regions using a machine learning approach, a generally increased activity, but decreased connectivity within a temporo-occipital theta band network was revealed during the reconfiguration of perception-action associations during inhibitory control. Additionally, a fronto-occipital alpha-theta interplay yielded a decrease in directed connectivity during reconfiguration processes, which was associated with lower error rates in behavior. Thus, adaptive behavior relies on both local increases and decreases of activity depending on the frequency band, and concomitant decreases in communication between frontal and sensory cortices. The findings reframe common conceptualizations about how adaptive behavior is supported by neural processes.