ChatGPT for Educators: Plan Lessons and Review Materials

ChatGPT can help educators plan lessons, rewrite explanations for different reading levels, and produce a first draft of quizzes, rubrics, and handouts. The key constraint is that it should never receive student names, grades, IEPs, or any other identifying information — that data is protected under FERPA and similar state laws.

Most of the practical value comes before class: planning and preparing materials, not grading or evaluating individual students. Tools like fenbs let teachers keep the prompts that work and the lessons learned from using AI in the classroom, so each new unit starts from a stronger foundation.

What educators use ChatGPT for most

The most common and lowest-risk uses are in lesson preparation. ChatGPT does not need to know your students to help you plan a lesson. It needs to know the topic, the grade level or prior knowledge you are assuming, and what you want students to be able to do at the end.

• Drafting lesson plans and daily objectives from a curriculum standard or unit goal.
• Rewriting an explanation at a simpler or more advanced reading level.
• Generating discussion questions for a text, concept, or event.
• Creating quiz questions, multiple-choice items, or short-answer prompts.
• Writing a rubric outline for a project or essay.
• Producing an extension activity or a challenge problem for students who finish early.
• Summarizing a long article or primary source into a student-readable version.

Planning a lesson with ChatGPT

Start with the outcome, not the activity. Telling ChatGPT the standard or goal you are working toward ("Students should understand how photosynthesis converts light into chemical energy, grade 9 biology") produces more useful results than asking it to "make a fun activity." The specifics of your students come from you; the draft scaffolding comes from the model.

Ask for a draft, then revise it. ChatGPT will produce a complete-looking lesson plan, but it does not know your pacing constraints, the prerequisite lessons, or the particular misconceptions your students often have. Your review of the draft is what makes it useful. Treat the output as a starting structure, not a finished product. The draft also gives you something concrete to evaluate: it is easier to spot what a lesson plan is missing when you can see what is there than when you are building from nothing.

A prompt that consistently produces a good draft is worth saving. Prompts that include the grade level, the prior knowledge you are assuming, the format you want (direct instruction, Socratic questioning, collaborative inquiry), and the time available tend to give better results than short one-line requests.

Adapting explanations for different levels

One of the most practical uses for teachers is rewriting. You have an explanation that works for most students and you need a simpler version for students who are struggling and a more technical version for students who are ready to go further. ChatGPT can produce both drafts from your original text or a description you provide.

Specify the reading level or the vocabulary assumptions explicitly. "Rewrite this at a 6th-grade reading level, avoiding the words 'osmosis' and 'concentration gradient'" gives better results than "make it simpler." Review the result for accuracy before using it — simplification can sometimes lose a critical distinction.

Reviewing AI-generated materials before class

AI-generated classroom materials need a teacher's review before use. ChatGPT can produce a quiz question with a plausible-looking but incorrect answer key, a historical example with a wrong date, or a science explanation that oversimplifies in a way that creates a misconception. These errors are not rare edge cases; they happen regularly.

Build the review into your workflow rather than treating it as optional. Check quiz answers against your textbook or a primary source. Read explanations for accuracy and for alignment with how your curriculum presents the concept. Note the errors you find so you can improve the prompts you use for that type of content. A few minutes of review before class prevents a corrected misconception from reaching students in the first place.

Keeping a record of the errors that appeared in ChatGPT-generated materials — and the prompts that produced better results after you adjusted them — is a practical teaching lesson. That record is more valuable as a shared note than as something buried in one teacher's chat history.

What educators must not do with ChatGPT

Student records, names, grades, disciplinary notes, disability accommodations, and any other personally identifiable information are protected under FERPA↗ (opens in a new tab). Do not enter any of that information into a consumer AI tool. This applies to summaries as well: "a student in my third period class who is struggling with fractions and has a 504 plan" is still identifying information in context.

Do not use ChatGPT to assess individual students or generate performance evaluations. Professional judgments about a student's progress and needs require the teacher's knowledge of that student and their professional training. An AI tool does not have that knowledge and is not a substitute for it.

Sharing materials and prompts with colleagues

A lesson plan that took twenty minutes to get right with AI is worth passing to a colleague rather than having them discover the same prompt adjustments independently. The same applies to the errors you found: a note that says "ChatGPT consistently gets the dates wrong for the Missouri Compromise — always check against the textbook" saves the next teacher the same surprise.

Assessment design: quizzes, rubrics, and feedback structures

Assessment design is one of the most time-consuming parts of teaching, and ChatGPT can accelerate the drafting stage significantly. For quizzes, describe the concept you are testing and the cognitive level you want: recall, application, or analysis. Ask for multiple-choice questions with a specific number of distractors, or for short-answer prompts that require explanation rather than just a definition. Then check every answer key against your course materials.

Rubrics follow a similar pattern. Describe the assignment type, the grade level, and the criteria you want to assess. ChatGPT can produce a rubric grid with performance levels and descriptors. Review it against the actual demands of the assignment. A rubric that uses language students find ambiguous or criteria that do not match what you actually value in the work needs revision before use.

Feedback comments are another useful application. If you have a set of common errors your students make on a particular type of assignment, describe those patterns and ask ChatGPT to draft feedback comment templates you can adapt. You fill in the specific observation for each student; the template gives you a consistent, clear starting point. Do not use student names or other identifying details in the prompt.

For project-based and creative assignments, ChatGPT can help you generate a range of example responses at different quality levels, which you can use to calibrate your own evaluation standards before you grade. Producing your own examples for anchor papers is a common assessment design practice; ChatGPT can give you a draft set to review and refine rather than writing them entirely from scratch. A rubric developed alongside example responses at each performance level is more consistent to apply than one written in the abstract, because the language maps to what the work actually looks like.

How to do this with fenbs

1. Create a free account. Start with a personal workspace; move to a team workspace when colleagues want shared access to the same prompts and lesson notes.
2. Create a project for a course or unit. Add a short description of the course level, the curriculum standards you are working toward, and any standing rules (never use student names, always check dates).
3. Add your best prompts in Skills and prompts. For each one, note the grade level, format, and what it reliably produces. Flag any known limitations.
4. Record errors and improvements in Lessons learned. When a prompt produced a wrong answer, write what happened and what adjustment fixed it.
5. Connect your AI assistant over MCP from Integrations and ask it to read the project context before helping with a new unit. It will start from your approved prompts and your lessons, not from scratch.

fenbs does not access your school's student information system or gradebook. It stores your project notes, prompts, and lessons. Keep all student data in your district's approved systems.