AI Literacy Index
A plain-language framework for what learners and teachers need to know, question, and direct when they work with AI agents.
Literacy for agents, not for slogans
The AI Literacy Index is NexEd's framework for talking about agents in school. An agent is not a search box. It can plan, call tools, and act across several steps. Students and teachers need language for that difference: what the agent is trying to do, which tools it used, where it may be fluent and wrong, and who remains responsible for the idea.
The Index is a vocabulary for curriculum and professional learning. It is not a ranked score of schools or students.
Know, question, direct
Knowing includes how an agent produces a plan, where a tool call can smuggle in an unsupported claim, and what memory a classroom agent retains. Questioning includes checking a step, noticing a missing source, and deciding when not to delegate the task. Directing includes setting the goal, the allowed tools, and the point at which a person must take the work back.
A strand is teachable when a student can point to it in their own session. They can say what the agent did, what they checked, and what they still own. If the only evidence is that a tool was opened, the strand was not taught.
How it stays honest
The Index does not publish a league table. When the lab reports what a strand looks like in a classroom, that reporting belongs in the journal, with sources a reader can open.