Schoolwork figure: Pew Research Center, 2026, surveyed September–October 2025.
Everyone here wants to run AI projects with students. Pick the scale you are working at.
Create the task, share one link, and students join without accounts. You see how each of them worked with the AI, in the same lesson.
Give a department or a whole staff one way of running AI projects. You get a single view across classes, and controls an administrator sets.
Everyone bringing AI into learning faces the same question: "How do I know real learning is happening — or is AI just doing the work?"
Every question asked. Every revision made. Every decision taken. It lands on one timeline while the lesson is still running.
See exactly where a student leaned on AI, where they pushed back, and where the thinking was their own.
Open the record with the student and ask how they got there. Not to catch them — to teach them.
Four ways the work students do with AI becomes something you can teach from.
Read a session back in the order it happened, not as a finished text.
See which parts of a draft the student made theirs, so your feedback lands on the writing they actually did.
Prompting is a skill you can teach. See who probes and who settles for the first answer.
Export the whole session — for a report, a progress meeting, or the student's own portfolio.
No rollout, no rostering, no training day. One task and one link.
A teacher creates a task and shares one link. Students join instantly — no accounts to provision, no roster to import.
Every prompt, AI response, revision and rejection lands on one timeline. Safety is calibrated by age band, in 7 languages and counting.
The record travels with the work: how much of the final text is the student's, where the AI carried them, and where they pushed back.