Direct questions

1. How do I choose an AI mentor for my child?

Choose the named person against three separate requirements: safe working arrangements, appropriate teaching, and demonstrable technical competence. Do not let a prestigious résumé or impressive project compensate for an unresolved safety issue.

Stanford’s National Student Support Accelerator identifies tutor selection, tutor consistency and monitoring learning as important components of tutoring quality. UNICEF’s AI guidance separately emphasizes children’s safety, privacy, inclusion and agency; neither framework establishes that a particular commercial mentor is suitable. (nssa.stanford.edu)

Use this proposed buying sequence:

  1. Write a learner brief. Record age band, interests, tasks the child can already complete, available time and the intended outcome.
  2. Resolve the non-negotiables. Establish identity, screening, parent-visible communication, account eligibility and data handling.
  3. Meet the actual instructor. Do not accept a sales consultation as a substitute for a teaching observation.
  4. Buy a bounded trial. Agree one small objective and an exit point before a substantial package.
  5. Review learner evidence. Ask what the child can now explain, change or debug with less help.

For each criterion, distinguish supplier assertion, document-supported evidence and observed practice. An agency’s written reporting policy belongs in the second category; a useful report about your child belongs in the third.

A good final interview question is: “At the end of the first block, what should my child be able to do that they cannot do now—and how will you show us?”

Sources cited on this page

  1. Tutoring Quality Standards | National Student Support Accelerator ↗nssa.stanford.edu