CHAILD - Children's Agency In the age of AI: Leveraging InterDisciplinarity

In today's rapidly changing digital world, children's lives are intricately intertwined with Artificial Intelligence (AI) systems, permeating their experiences through various platforms like connected toys, apps, voice assistants, and online learning platforms. While these systems offer opportunities for supporting children's development and learning, they present grave risks to children's autonomy and agency through the pervasive practice of datafication and manipulation of their digital experiences and behaviours.

Our project aims to address the pressing issue of fostering children's digital autonomy and resilience by promoting what we call 'algorithmic agency' in children, referring to children's ability to make independent choices and take actions within the digital environments governed by algorithms. The current lack of focus on supporting children's algorithmic agency greatly hinders the critical development of their digital resilience. Agency forms the basis of individuals' ability to regulate their thoughts, motivations, and behaviours based on their beliefs about themselves. However, currently, children's agency is particularly vulnerable to technological influence, designed to exploit their autonomy and shape their behaviours and decision-making. The notable absence of a shared, interdisciplinary understanding of children's agency in the digital context prevents the development of effective approaches to protect and nurture children's agentive capacities of self-regulation, self-reflection, intentionality, and foresight throughout their datafied childhood.

In this project, we aim to leverage the interdisciplinary expertise and knowledge of computer scientists, learning scientists, social scientists, philosophers, to develop a comprehensive understanding of what algorithmic agency means for children in the digital context and define effective methods to create AI systems that actively foster and promote children's agency while respecting their values and rights. To achieve this, we will not only work directly with children and understand their voices but also with diverse groups of stakeholders, including parents, teachers, developers, and policymakers.

Our objectives are organised around four themes: (i) a foundational understanding of children's algorithmic agency; (ii) new methods for building agency-fostering AI systems for children, (iii) timely guidance for parents, educators, and developers, and (iv) evidence-based AI policies that prioritise children's needs.

The project brings together an experienced team of computer scientists, learning scientists, social scientists, philosophers, and public policy partners to delve into the complexities of algorithmic agency for children. Through its interdisciplinary insights, the project aims to deliver a deep understanding of the nature of children's agency and their associated needs in the context of their use of AI-enabled products and interventions. We will co-design principles for creating agency-fostering child-centric AI systems, and produce breakthrough empirical evidence through iterative prototype evaluation. We will create lasting impact by raising awareness of children's algorithmic agency for the public, industry, and policymaking.

The outcomes of our project will provide timely impact and insights that bear on regulatory development, social welfare, and responsible innovation. It will shift the current mindset where children are passive consumers of technologies towards one where AI systems are designed to respect, nurture and scaffold the development of agentive competency in children, and equip them with better digital resilience and competence.