Dating back to the seminal studies of Sherif (1935), there is robust evidence that social norm learning is able to shape perceptual decision making in a persistent manner. But what mechanisms underlie this effect? Here, we propose the new attentional alignment hypothesis. According to this hypothesis, norm learning alters feature-based visual attention. In particular, we hypothesize that norm learning increases visual attention toward norm-congruent relative to norm-incongruent stimulus features. Using steady-state visual evoked potentials, our results show for the first time that norm learning can persistently alter early attentional processes in the visual cortex. As predicted by the attentional alignment hypothesis, individuals' feature-based attention was tuned toward norm-congruent and away from norm-incongruent features. This bias persisted even when norm information was no longer available. (PsycInfo Database Record (c) 2023 APA, all rights reserved).