Teacher AI competency is not measured by how many tools someone has tried. It is the ability to choose an appropriate task, produce a useful draft, recognize risk, verify the result, protect learners, and remain accountable for the final teaching decision.

1. Define the teaching purpose

Can the teacher state the learning objective, the learner group, and why AI is appropriate for this task? Competent use begins by deciding what should remain human judgment and what can be accelerated safely. “Create something engaging” is not yet an instructional purpose.

2. Give constrained, useful instructions

A strong request includes age, language, objective, source boundaries, format, difficulty, accessibility needs, and desired feedback. The teacher should be able to revise one variable at a time and ask the system to identify uncertainty instead of inventing missing facts.

3. Verify content and assessment quality

The teacher checks factual accuracy, calculations, accepted answers, distractors, explanations, reading level, curriculum fit, and cultural assumptions. They preview the learner experience and solve every scored item. Fluency, confidence, and polished formatting are never substitutes for verification.

4. Protect privacy, ownership, and access

Competent use follows school policy and avoids sending names, identifiers, sensitive work, or confidential records to unapproved services. The teacher checks reuse rights, attributes sources when required, and ensures that captions, alternatives, contrast, keyboard access, and language do not exclude learners.

5. Use AI transparently in the classroom

The teacher can explain where AI contributed, what was reviewed, and what students are expected to do independently. They design for agency rather than surveillance, offer a non-AI route when appropriate, and do not use generated output alone for grading, discipline, placement, or diagnosis.

6. Learn from evidence and document decisions

After use, the teacher compares the activity with learner work, discussion, errors, access issues, and feedback. They record important prompt changes and corrections, share reusable lessons with colleagues, and stop using a workflow when its time saving does not justify its risk or review cost.

  • Purpose is explicit.
  • Instructions are constrained.
  • Every claim and scored item is checked.
  • Privacy and accessibility are protected.
  • Responsibility remains with people.