ABSTRACT
Children and youth are surrounded by AI in many of the products they use in their daily lives, from social media to education technology, video games, smart toys, and voice assistants. AI can affect the videos children watch online, their curriculum as they learn, and the way they play, and interact with others. For the purposes of this chapter, we adopt UNICEF’s definition of AI. AI can be used to facilitate children’s development and empower them in their activities, but children are also vulnerable to potential risks posed by AI, including bias, cybersecurity, data protection and privacy, and lack of accessibility. AI must be designed to respect the rights of the child user and to provide equal opportunities for all children. Several organizations, including UNICEF and the World Economic Forum (WEF), have developed guidelines for child-centric AI, but the challenge of designing and implementing responsible and trusted child-centered AI remains complex to address. In this chapter, we describe challenges and propose a regulatory sandbox environment as a means to address children, businesses, regulations, and society as a whole. For the rest of the chapter, whenever we use the term “sandboxes,” we refer to “regulatory sandboxes.”
