Authenticity and co-design: On responsibly creating relational robots for children

Milo Phillips-Brown, Marion Boulicault, Jacqueline Kory-Westland, Stephanie Nguyen, Cynthia Breazeal

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract / Description of output

Meet Tega. Blue, fluffy, and AI-enabled, Tega is a relational robot: a robot designed to form relationships with humans. Created to aid in early childhood education, Tega talks with children, plays educational games with them, solves puzzles, and helps in creative activities like making up stories and drawing. Children are drawn to Tega, describing him as a friend, and attributing thoughts and feelings to him ("he's kind," "if you just left him here and nobody came to play with him, he might be sad"). Scholars and members of the public alike have raised the alarm about relational robots, worrying that the relationships that people, and especially children, form with such robots are objectionably inauthentic. We, members of an interdisciplinary team that developed and studies Tega, make this inauthenticity worry precise and offer practical recommendations for addressing it. We distinguish two kinds of (in)authenticity -- inauthenticity as unreality, and inauthenticity as deception -- arguing that neither is just what others have thought it is. With our distinction in hand, we argue that the authenticity concern can be met only through co-design methods -- methods that give stakeholders of all kinds (e.g. parents, educators, and children themselves) a genuine say in how relational robots are built.
Original languageEnglish
Title of host publicationAlgorithmic Rights and Protections for Children
EditorsMizuko Ito, Remy Cross, Karthik Dinakar, Candice Odgers
PublisherMIT Press
Number of pages37
ISBN (Electronic)9780262374323, 9780262374316
ISBN (Print)9780262545488
Publication statusPublished - 2023


Dive into the research topics of 'Authenticity and co-design: On responsibly creating relational robots for children'. Together they form a unique fingerprint.

Cite this