UNICEF's Policy Guidance on AI for Children offers recommendations for crafting AI policies and systems that safeguard child rights. Despite its potential for innovation, AI also poses risks to children's privacy, safety, and security, which is often overlooked in existing AI policies. As part of UNICEF's AI for Children project, the Policy Guidance on AI for Children aims to ensure the protection of children's rights within government and private sector AI policies and practices while increasing awareness of how AI systems can impact these rights. The guidance explores AI systems in relation to their effects on children and was initially released as a draft in September 2020. Input from various stakeholders, including AI professionals, companies, government agencies, civil society, international organizations, academics, and both adult and child citizens, was solicited and integrated into version 2.0, published in November 2021. To support the implementation of the guidance, UNICEF has made available a list of online resources and a set of practical implementation tools.