Artificial intelligence is playing an increasingly central role in all our lives, for children as much as adults. Children now interact with AI through toys and games, the internet and smartphones, and software at school.
But despite the fact that children are growing up in a society that is shaped by AI, there is currently little consultation with them to find out how they are affected by this technology. The Turing highlighted this issue in its recent response to the government’s statement regarding the use of generative AI in education.
Since 2022, our ‘Exploring children’s rights and AI’ project in the Turing’s public policy programme has sought to identify the ways in which children understand, experience and engage with AI. The project aims to provide much-needed guidance for policy makers, AI developers and affected stakeholders to inform future approaches to child-centred AI and contribute to policy and research that positively impacts children in the UK and across the globe. An overarching goal of the project is to promote children as a key stakeholder group in the ethical development of AI systems.
“I am excited to learn more [about AI] because I am not sure about it; I think it is on my phone. I want to learn more about the inner workings of it.” – Project participant, age 8, Shetland
We worked with Children’s Parliament and the Scottish AI Alliance to engage 87 children age 7-11 across Scotland in thinking about AI. From September 2022 to February 2023, we ran in-person and online workshops with schools in Edinburgh, Glasgow, Stirlingshire and Shetland and developed “missions” for the children, from an online game to help them learn about machine learning, to the “Playground Predictor” – an exercise encouraging them to think like an AI system to explore some of the problems of AI bias. The missions helped us to increase the children’s understanding of AI, engage with them about their rights and AI and, most importantly, learn about what matters to them in this domain.
The children shared their thoughts on bias, fairness, AI applications and how AI may be better developed to uphold children’s rights. They quickly grasped difficult concepts like how algorithms are used in prediction, recommendation and classification. Something that came through particularly strongly was the children’s desire to be brought into discussions around new technologies, so that their perspectives could be considered.
“I didn’t know before that AI could be used in healthcare to help people.” – Project participant (Primary 5), Stirlingshire
Recognising that AI systems inherit biases in the data they are trained on, our participants wanted to ensure that children from many different backgrounds would be consulted when data is gathered for new systems, to make sure that the outputs don’t privilege a particular group. The children felt that they should be consulted on data collection, as well as the ways in which AI might impact their families and communities – they shared worries that if AI negatively affected the adults who cared for them, then children would be impacted too.
The children felt strongly that they should be given opportunities to learn about how AI systems are created and trained, so that they can be empowered to participate in conversations about how AI impacts children’s rights, and help to generate creative ideas for new, better systems. The children also wanted to see more young people involved in AI education and decision-making across age groups, and they were excited to explore what meaningful participation in AI development and governance could look like in the future. This will be a focus of phase two of our project, which will continue to engage with children across Scotland through collaboration with Children’s Parliament and the Scottish AI Alliance.
Alongside this work, our team is examining the legal underpinnings of child-centred AI, and we will soon be publishing a review of existing children’s rights and digital technologies frameworks, as well as an in-depth analysis of the various themes and focus areas of transnational organisations’ frameworks. Other ongoing work includes a collaboration with the Council of Europe to explore member states’ legal frameworks addressing children’s rights and AI.
This programme of work aims to substantially inform and enable future best practice in child-centred AI both across the UK and internationally, ensuring that children’s rights and children’s voices are at the heart of AI development and governance.
Read our report:
Exploring Children’s Rights and AI
Images: Roberto Ricciuti
