Before you continue

This is a research demonstration for adults. It shows, side by side, how a general AI response differs from one conditioned on a declared developmental stage. It is not a product, and it is not for children to use.

By continuing you confirm you are 18 or older and understand this is an illustrative demonstration of why AI products and EdTech used by people under 18 need a framework like this.

No prompts are stored. This demonstration does not log or retain what you type.

Baseline

A developmental framework for minors interacting with AI.

Type a prompt. The left shows a general AI response; the right shows the same prompt conditioned on a declared developmental band, with the framework's decisions made visible. Band is declared, never inferred.

What this framework is grounded in

Baseline draws on established bodies of work across several domains:

  • Developmental capacity — cognitive-stage models (concrete and formal operations), theory-of-mind development, and literacy-development stages (the shift from learning to read to reading to learn). Grade boundaries approximate transitions across these.
  • Theory of mind & anthropomorphism — how children attribute mind and feeling to responsive systems, and attachment research.
  • Content-harm & age-appropriate design — regulatory frameworks defining content withheld from minors, and the under-13 / teen distinctions.
  • Reading level & complexity — established text-complexity and readability measures used as per-band acceptance checks.
  • Routing & safeguarding — trauma-informed, trusted-adult routing practice from child-safety organizations.

On grade bands as a proxy. Grade is an imperfect proxy for developmental stage; individual children vary, and grade placement does not guarantee a given stage of cognitive, social, or literacy development. The framework bands by grade because it is the declarable, institutionally-known variable, consistent with the principle that a child's band is declared by the institution rather than inferred by the model. Banding by developmental stage directly would require individual developmental assessment that is neither available nor scalable in a school setting. The bands are deliberately coarse to accommodate within-grade variation, and grade boundaries are chosen to approximate well-established developmental transitions rather than to assert sharp cutoffs.

A full annotated bibliography is in development and available on request.

Have we missed something? This framework is a work in progress. Suggest a source, a domain, or a gap.

Email this suggestion
Feedback welcome

This is a work in progress and I'd value your read. Where did it fail, in your view? Where did it get something right? What's missing, wrong, or worth pushing on? Candid reactions are the most useful.

Send your thoughts. What worked, what didn't, where it broke.

Email this feedback