ChatGPT for K-12 Teachers: Lesson Planning and Student Feedback

ChatGPT helps K-12 teachers with the written preparation that takes the most time outside the classroom: drafting lesson plans, generating differentiated activities for different learners, writing feedback templates, and creating assessment materials. The teacher reviews everything before it goes to students, applies professional judgment about what will work for their specific class, and never enters student names or personal data into any AI tool. Tools like fenbs keep teaching prompts and curriculum decisions in one place so each AI session starts from the agreed structure, not from a blank page.

The Family Educational Rights and Privacy Act (FERPA) protects the privacy of student education records. Any information that could identify a specific student should stay out of AI tools. Use generic descriptions like "a student reading two grade levels below" rather than any student's actual name or details.

Drafting lesson plans

Give ChatGPT the grade level, the subject, the standard or learning objective, the available time, and any relevant context about the class (for example, that the class includes English language learners). Ask it to draft a lesson plan with a hook, direct instruction, guided practice, and an exit ticket.

Review the draft against your curriculum map. ChatGPT does not know your pacing guide, the sequence you have followed, or what students already know. It will produce a structurally reasonable plan that may cover something the class has already mastered or skip a prerequisite. You fill that gap.

A lesson plan that works well once is worth saving as a template. Record the prompt that produced it, the adjustments you made, and the grade and standard it covers. The next time you need something similar, you start from a template rather than a new request.

Generating differentiated activities

Differentiation is one of the most time-intensive parts of lesson preparation. Ask ChatGPT to produce three versions of the same activity at different complexity levels: one for students who need more support, one at grade level, and one for students who need extension. Describe the activity and the learning goal first, then ask for the variations.

Check each version for accuracy and appropriateness. A simplified version that removes important content rather than scaffolding it is not useful differentiation. An extension that requires background knowledge the class has not covered is equally problematic.

Keep descriptions generic. "Students who are reading below grade level" or "students who need an additional challenge" is specific enough for the AI to generate useful variations without identifying anyone.

Writing feedback templates

Feedback templates help you give consistent, useful written comments on student work without writing each one from scratch. Ask ChatGPT to produce a set of comment starters for a specific assignment type or skill area. For example: "Give me ten sentence starters for written feedback on argument structure in a fifth-grade persuasive essay."

Use the templates as starting points. Specific feedback is more useful than generic praise, so adapt the template to what you actually observed in a student's work. The template handles the framing; you add the observation.

Never include a student's name, grades, or specific performance data in a prompt to ChatGPT. If you want feedback on a writing sample, describe the writing pattern generically: "a student who consistently omits evidence sentences in the body paragraphs."

Creating assessment materials

ChatGPT can draft quiz questions, exit tickets, formative assessment prompts, and rubrics. Give it the standard or skill, the grade level, and the format you need. Always check the content for accuracy, age-appropriateness, and alignment with what you have taught. ChatGPT does not know your curriculum sequence.

For multiple-choice questions, check that the correct answer is unambiguously correct and that the distractors reflect plausible misconceptions rather than obvious wrong answers. AI-generated distractors are sometimes too easy to eliminate.

Keeping curriculum decisions consistent

A department or grade team often has decisions about curriculum that should carry through every lesson plan: the writing format used for essays, the vocabulary instruction approach, the standard feedback rubric. If those decisions are only in someone's head or in a document no one reads, an AI assistant will produce output that ignores them.

Record agreed curriculum choices as rules in a shared project so every assistant any teacher connects reads them. "All essay feedback uses the department rubric headers" is a rule that should appear in every lesson planning session, not just when someone remembers to mention it.

What ChatGPT cannot do for teachers

ChatGPT cannot observe your class, know your students, or make professional teaching decisions. It does not know your school's policies, your district's approved curriculum, or the specific standards your state requires. Treat its output as a draft that your professional judgment approves, modifies, or rejects.

Any AI tool used with student data must comply with FERPA and your district's data privacy policies. Check with your district before using any AI tool in ways that could involve student information. The US Department of Education↗ (opens in a new tab) publishes FERPA guidance.

How to do this with fenbs

1. Create a free account and add a curriculum project for the grade or subject you teach.
2. Save the lesson plan prompts, differentiation formats, and feedback templates that work well in Skills and prompts.
3. Record agreed curriculum choices as Decisions and rules: essay structure, rubric format, vocabulary approach, department policies.
4. Connect your AI app over MCP from Integrations so it reads the curriculum rules and prompts at the start of each planning session.
5. Never enter student names, grades, or personal data into any prompt. Describe learning needs generically.