The toys of the future may take many different forms. As artificial intelligence becomes part of children’s play and social interaction, it becomes increasingly important to find methods for involving children in the development of those toys. Above is an example of how speculative design can support that process. It may look rather unlikely, admittedly, but this prototype has provided new insights into how children think about interacting with AI in play.
When AI enters the playroom, children must become co-designers of the future
There is a classic paradox that asks what happens when an unstoppable force meets an immovable object. As artificial intelligence makes an almost unstoppable entrance into every layer of children’s play and everyday lives, it may be fitting to reformulate the question: what might happen when a system-driven language model meets the unspoken nature of play.
The question matters because it touches on something fundamental in how we develop and engage with the world when free imagination is shaped by fixed AI formulas and when toys can influence children’s play and upbringing.
The frightening scenario in this development is easy to imagine: over time, we may lose our ability to form human relationships in the ways we once did, and children may lose the vital capacity to invent and create play and togetherness without a script – a crucial human ability. The first steps towards a dream scenario may, at least for now, look rather implausible.
Kolding School of Design has brought together Danish partners in a research collaboration with Yonsei University in South Korea. The aim is to take the first steps towards creating a balance between technology and free imagination, and to establish an international network that in the future can collaborate on shaping the role of technology in the world of play with a broader perspective than before.
- Artificial intelligence is changing the way children play and learn. If AI for children is developed purely from a technical perspective, the future will become narrow and driven by technology and market logic. That is unlikely to be desirable, yet we do not know how children themselves perceive their relationship with new AI-based toys. Our project can create an understanding of what we are about to introduce into the playroom and invite children to contribute to the work. That insight has not existed before, says project lead Jesper Falck Legaard from Kolding School of Design’s LAB for Play Design.
The project began in 2025, and since then the research group has moved into the field known as speculative design. The project is called PL_AI, where the speculative approach invites participants, using simple materials, free associations and playful methods, to imagine how AI might become an active part of children’s play worlds.
Associate Professor Jesper Falck Legaard from LAB for Play Design is the project lead for PL_AI, which is building a network across disciplines and national borders to develop methods for involving both children and developers in shaping the toys of the future.
Around the world, researchers are investigating how artificial intelligence may influence children’s development. What makes the Danish-led project distinctive is the composition of the group, combining South Korean technological development with Danish design thinking and a strong understanding of play in practice.
One of the private partners in the project is Martin Rauff-Nielsen. Throughout his career he has worked with creating frameworks for play, developing methods that encourage people to play together – all of us. For a period he did so on a national scale when he appeared as Game Master Martin on Danish Broadcasting Corporation (DR), where play, imagination and open storytelling became shared experiences in the fantasy world of Barda from 2006 to 2013.
Today he runs his own company, Baaring Stories, and has a deep understanding of the nature of play and of the unspoken dynamics that shape shared storytelling. That insight is what he brings into the project, which in return provides new knowledge and networks for future work.
- AI works best when language is concrete and unambiguous, but play thrives on subtle undertones, unwritten rules and the right to say yes or no – things a programme will never fully capture. This field is incredibly relevant to explore through speculative design, because these prototypes allow us to frame a possible future. The approach is hugely valuable for the project and for me personally, because designers, engineers, philosophers, business people and others can discuss, challenge and concretise the conversation about what AI is – and what it is not.
- With PL_AI we have an opportunity to shape the framework for artificial intelligence so that children in the future do not end up with artificial play. When a teddy bear can answer back, it is no longer just a toy. It can influence a child’s understanding of friendship, authority and trust. That is why it makes sense to create a space where we can pause and explore these questions before technological development has already defined what this technology will be in children’s lives.
Martin Rauff-Nielsen has worked with creating frameworks for play throughout his entire career and has over the years developed methods that encourage us to play together.
The prototypes proved their value back in Denmark, when 24 pupils from a Year 3 class were invited to build their own versions of future AI toys together with researchers from Kolding School of Design. The exercise provided insight into children’s perceptions of artificial intelligence.
Jesper Falck Legaard explains that one group developed a helper that could remind them to brush their teeth – much like a parent would. Another group built a bed for their AI device, because their artificial friend would obviously need somewhere to sleep. The latter in particular prompted reflection within the research group:
- We are not asking children to explain their opinions about AI. We ask them to imagine what AI might look like for them. Our task is to observe what happens while they build. When they started building a bed for the technology, we could see how quickly the relationship becomes personal. That tells us something new about trust and expectations. It also tells us that the next step is to dive even deeper into the layers surrounding ethical questions and guidelines, which we will continue discussing with experts in the field, says Jesper Falck Legaard.
The project is funded through a grant from the Danish Agency for Higher Education and Science under the GINP – Global Innovation Network Programme. GINP supports networking activities between Danish and international research and innovation partners. The next step for PL_AI will be another visit to Yonsei University to gather and analyse experiences and insights, and to strengthen professional relations between the partners. The project will conclude in 2027 with a website where results and experiences will be shared.
