A primary method of collecting evidence against a group of misbehaving agents (e.g., a cartel) is to offer whistleblower rewards to agents who come forward with evidence. But even before a whistleblower comes forward, the regulator typically possesses her own private evidence, though it may not guarantee conviction in court. This paper studies how revealing this private evidence to group members can facilitate whistleblowing. The more damning the regulator's evidence, the more incentivized, all else equal, a whistleblower is to come forward. I formalize this environment using information design in games. A principal (sender) informs two agents (receivers), about a state (the evidence state) that affects the payoffs to whistleblowing. I first ask, if agents can communicate amongst themselves, can the principal improve over public communication with the group? I answer the question in the affirmative, and provide conditions under which the principal's value is independent of whether agents can communicate or not. Second, interpreting the evidence state as the probability of conviction without a whistleblower, I characterize optimal outcomes and show that the likelihood of whistleblowing is increasing in this probability. I show that the principal can facilitate more whistleblowing in groups with more asymmetrically distributed gains from misbehavior across agents. Finally, I show that there are simple information structures that, though potentially sub-optimal, robustly improve over public communication.