While Artificial Intelligence enables substantial productivity gains from delegating tasks to machines, it may facilitate the delegation of unethical behaviour. This risk is highly relevant amid the rapid rise of 'agentic' AI systems. Here we explore whether delegating tasks to machines might increase human immoral behavior. We also study the factors that may influence this effect. Factors range from how we delegate to machines (e.g. using natural language vs. rule-based programming) to whether some AI systems are more compliant than others when asked to perform unethical tasks. We explore design and policy strategies to mitigate the ethical risks from delegation to machines. These include guardrails that operate both on the humans and the machines.