Minimax is a decision rule in Game Theory for minimizing the maximum possible loss in a worst-case scenario. It assumes an opponent plays optimally to maximize their own gain. This strategy is fundamental in Artificial Intelligence for two-player games.