We analyze maximal cartel prices in infinitely-repeated oligopoly models under leniency where fines are linked to illegal gains, as often outlined in existing antitrust regulation, and detection probabilities depend on the degree of collusion. We introduce cartel culture that describes how likely cartels persist after each conviction. Our analysis disentangles the effects of traditional antitrust regulation, leniency, and cartel strategies. Without rewards to the strictly-first reporter, leniency cannot reduce maximal cartel prices below those under traditional regulation. Moreover, in order to avoid adverse effects fine reductions should be moderate in case of multiple reporters. Our results extend the current literature and partially support existing leniency programs.