The most common way companies seek to improve legal rule knowledge is by adopting corporate policies and communicating them to their employees. While expected by multiple regulatory authorities and widely employed by corporations –by now also in shorter formats– evidence documenting the effectiveness of such corporate policies in improving legal rule knowledge, particularly from the field, is lacking. We thus conducted an extensive online field experiment in which we randomly assigned 1,235 employees of an international technology company to four treatments. In the Long Policy treatment, employees received a 19-page traditional policy written in standard legalese language. The Short Policy treatment entailed a shortened 4- page version of this classic policy, and in the Infographic treatment, employees received an illustrated overview of the essential rules on a single page. Notably, a fourth group received no policy and served as a Control treatment. After reading the policy, employees completed several measures assessing their rule knowledge, perceived social norms and played an incentivized bribery game. We find that a) none of the types of policies presented improves rule knowledge or reduces corrupt behavior compared to the control treatment, and b) no differences exist between the policies in influencing rules knowledge or in reducing corrupt behavior. Instead, we find robust evidence indicating that people form their beliefs about corporate rules and decide to engage in corrupt behavior based on what they consider to be the norm.