A healthcare provider is considering integrating artificial intelligence (AI) to predict patient treatment outcomes. What ethical principle should guide them in ensuring that the system does not cause unintended harm to patients?