Informing Age Appropriate AI Systems for Children - CHI'2022 Honourable Mention

AI systems are becoming increasingly pervasive within children's devices, apps, and services. However, it is not yet well-understood how risks and ethical considerations of AI relate to children. To this end, in late 2021, we conducted a review of leading general AI frameworks and those for children, and a systematic literature review of 188 papers about children's AI systems. Our analysis revealed a wide assortment of application domains in children's lives, including personalised tutoring/intervention, medical diagnosis, harms and safety, social robotics, personalised entertainment, public services, speech recognition, and emotion recognition, and age recognition.

More importantly, we noticed that most existing AI systems' designs addressed only a small subset of AI principles among those we identified. We synthesised our findings into a framework of "Code for Age-Appropriate AI", which aims to provide timely input to emerging policies and standards and inspire increased interactions between the AI and child-computer interaction communities.

This new code includes 5 principles:

  • Fairness, equality, inclusion and access, which is derived as a combination of the fairness and non-discrimination principles and that of universal inclusion, relating not only to discovering the needs of diverse groups but also ensuring that all children are treated fairly and equally.
  • Transparency and accountability, to identify a chain of responsibility for system (mis)-behaviours, and present them this information in an easy-to-understand manner. We particularly advocate the consideration of accountability and transparency from more than the system perspective.
  • Privacy and manipulation and exploitationto protect children's rights and safeguard them from being manipulated and exploited.
  • Safety and safeguarding, that is to ensure systems would not do harm to children, as well as protect children from harm.
  • Sustainability in a child-specific context should refer to supporting the long-term development of children, and in a way that considers the important aspect of meeting the developmental needs of children.

The full paper received Honourable Mention at ACM Conference on Human Factors in Computing Systems in April 2022.

[short video] [full presentation recording][[pdf]