Kids get just as creative with ChatGPT as adults do, but their brains coordinate less while doing it — is that the AI's fault?

This explores whether the weaker brain-network coordination seen in young children during AI co-creation comes from the AI, or whether it would show up in any creative task at that age. The short answer is that the corpus can't settle it yet. The one study here that looked at children's brains during creative work only measured them while they were working with ChatGPT, and it had no comparison task without AI. That missing baseline is the real gap, and it's worth knowing about before reading too much into the result.

Here is what the study did find. Six-to-seven-year-olds and adults came up with equally creative ideas while talking with ChatGPT, but the children's frontoparietal, salience and dorsal attention networks were much less connected to each other. Adults rely on these systems to plan, notice what matters and stay focused while generating ideas Why do children's brains show weaker control networks during AI dialogue?. The most likely reading is developmental: these control networks are still maturing in early childhood, so you would expect weaker integration on any demanding task, with or without AI. But because the study only compared children with adults, and never AI work with non-AI work, it can't tell 'young brains are less coordinated' apart from 'young brains coordinate less when AI is doing part of the work.'

Research on adults suggests the second possibility deserves to be taken seriously. A four-month EEG study found that brain connectivity went down as reliance on AI went up. People using an LLM showed the weakest neural engagement and had trouble remembering work they had just produced Does AI assistance weaken our brain's ability to think independently?. If adults' networks disengage when an LLM takes over part of the thinking, children's still-developing networks could be affected at least as much. Whether working without AI would build those networks up is exactly the experiment nobody in the collection has run.

The surprising detail is that the children's creative output matched the adults' while their brain activity looked very different. That fits a broader argument in the collection: AI separates the finished intellectual product from the thinking that normally produces it Does AI separate intellectual form from the thinking behind it?. For children, this means a good-looking result made with AI says little about whether the mental skills behind it are being exercised. That's why the missing comparison with non-AI tasks matters for anyone thinking about AI in early education.