Much of the conversation around AI for children still focuses too narrowly on the model itself. Is it smart enough? Safe enough? Accurate enough? These questions matter, but they are no longer sufficient. The next generation of child-facing AI will not be defined by better foundation models alone. It will be defined by better system design.
For children, an AI experience is never just the model. It's the model, the interface, the surrounding safeguards, the role of trusted adults, the context of use, and the product form itself. At FINH we believe the most useful way to think about child-facing AI is not as a model problem, but as a systems design problem. We call our evaluation toolkit AIMS+.
A is for Agent
Different use cases require different AI roles. A child may need a creative spark in one moment, a making copilot in another, a tutor in another, a reflective coach in another. A brainstorming partner should behave differently from a step-by-step guide. The first question is always: is this the right AI role for this child, in this context, for this task?
I is for Interface
Interface design shapes how children experience agency. A conversational assistant, a visual studio, a guided workflow, voice, touch, physical controls , each creates a different relationship. The same intelligence can feel empowering in one interface and overbearing in another. Interface design is a core part of intelligence design, not a wrapper placed around it afterwards.
M is for Mentors
We use the word mentors deliberately. "Human in the loop" is correct but too cold for children's lives. Parents, teachers, facilitators , they aren't only oversight mechanisms. They're conversation extenders, interpreters, and reviewers. The most powerful systems design trusted adults into the loop in lightweight, meaningful ways, not as blockers but as enablers.
S is for Shape
Shape is the overall structure and form of the experience , product form, delivery mode, degree of boundedness, software vs hardware vs hybrid. Shape also covers the rhythm: open-ended or guided, solitary or collaborative, fast and generative or slow and reflective, always available or intentionally bounded. These structural choices strongly influence what children actually do.
+ is for the system conditions
The plus matters because child-facing AI does not succeed on design alone. It depends on safety systems, age-appropriate orchestration, governance, moderation, parental controls, deployment environments, and the operational realities of running live products used by children and families. AIMS is the design lens; the plus is the responsible system around it.
Why creativity and metacognition sit at the centre
The most valuable AI experiences for children may not be the ones that generate the fastest output. They may be the ones that help children think more clearly about what they're making, why they're making it, and how to improve it. Metacognition is what turns creative activity into developmental value. When evaluating AI systems for children, we shouldn't just ask whether the child made something , we should ask whether the system helped the child think.