Handshake AI Safety Policy Evaluator at Handshake focuses on distinguishing between depictions of violence and enabling violence in AI model outputs. The role requires evaluating user requests and model responses to apply appropriate policy categories.