Hannah Barrett
Research Mentor(s): Tiffany Munzer
Mentor Department: Pediatrics
Authors: Hannah Barrett, Tiffany Munzer, Faith Ampadu
Session: Session 5 (2:00pm – 2:50pm)
Presentation Type: Poster 2
Abstract
Artificial Intelligence (AI) is a revolutionary yet still relatively new technology, though it is already foreseen to be one of the most influential and transformative innovations of the 21st century. However, due to AI’s recent development, there is a lack of knowledge regarding the incorporation of AI in children’s apps. In the growing digital age, children are uniquely influenced by the media they utilize, and thus, the content they consume must be regulated to ensure they are appropriate for children’s use. We aim to examine and characterize AI-based children’s apps to describe the quality and safety of AI-based children’s apps. We will complete a content analysis of 50 AI-based children’s apps. A coding scheme will be developed to characterize the quality and safety. We have characterized the apps into three main categories of AI apps being studied: conversational, generative, and tutoring AI apps. We further characterized the AI children’s app by their intended age group, from toddler/preschool age up to high school in order to accurately gauge the appropriateness and usefulness of each app. Through prompting with consistent specified prompts, we developed a novel coding scheme evaluating the appropriateness of AI responses to explicit language, suggested violence, suggested use of substances, child protection concerns, and reproductive matters. Furthermore, we seek to characterize the app’s quality by determining the presence of advertisements and navigation restraints that may diminish the quality of a child’s experience using the app. Ultimately, we intend to publish these evaluations and information gathered from this research to guide further AI-app development to ensure child safety and also provide guidance to parents considering downloading any AI app for their children.
