How Teachers Use AI to Plan Lessons: A Practical Workflow

Lesson planning eats evenings and weekends. Writing objectives, building activities, making 3 versions of a reading for different levels, then writing the quiz and the rubric can easily take longer than teaching the lesson itself.

AI tools can take a real share of that load. In a Gallup and Walton Family Foundation survey of 2,232 US public K-12 teachers conducted in spring 2025, teachers who used AI at least weekly estimated they saved 5.9 hours a week. The same survey found that only 32% of teachers used AI that often, and 40% did not use it at all.

The difference between a useful AI lesson and a sloppy one comes down to workflow. This guide gives you one you can use tomorrow.

What AI is good at in lesson planning

AI tools do best on the drafting and adapting tasks that take time but not deep judgment:

• Drafting learning objectives from a standard
• Generating lesson hooks, warm-ups and exit tickets
• Writing practice questions at several difficulty levels
• Rewriting a text at a different reading level
• Building rubrics from an assignment description
• Creating discussion prompts, sentence starters and vocabulary lists
• Suggesting accommodations and scaffolds

What AI does poorly: knowing your students, matching your local curriculum exactly, and getting every fact right. That is your job, and it is why the workflow below always ends with a human review.

Tools teachers use for AI lesson planning

You do not need a special tool to start, but education-specific tools save prompt-writing time.

Some of these are free for teachers. Khan Academy, for example, made Khanmigo for Teachers free for US teachers with funding from Microsoft, offering tools for rubrics, exit tickets, leveled texts and IEP support. If you want to build your own assistant, our guide on building a custom GPT for your team covers the setup.

Step 1: Draft with a specific prompt

Generic prompts produce generic lessons. Give the AI the same context you would give a student teacher.

A strong lesson-planning prompt includes:

1. Grade and subject
2. The standard (paste the full text, not just the code)
3. Time available
4. What students already know
5. The format you want (5E model, I do/we do/you do, stations)
6. Materials you have

Example prompt:

“Create a 45-minute 7th grade science lesson on the water cycle aligned to this standard: [paste standard]. Students already know the 3 states of matter. Use the 5E model. Include a 5-minute hook, a hands-on activity using cups, ice and plastic wrap, 3 discussion questions and a 3-question exit ticket. List materials.”

Then iterate. Ask for a different hook, a shorter activity or a version for a block schedule. Treat the first draft as raw material.

Step 2: Differentiate for your students

Differentiation is where AI saves the most time, because producing several versions of the same material by hand is slow.

Useful prompts:

• “Rewrite this passage at a 4th grade reading level while keeping the key vocabulary: [paste].”
• “Create 3 versions of this worksheet: one with sentence starters and a word bank, one standard, one with an extension challenge.”
• “Suggest scaffolds for English learners at an intermediate level for this activity.”
• “Turn this lesson into a version that works for a student who needs movement breaks.”

For students with learning differences such as dyslexia, combine leveled texts with built-in tools like text-to-speech. Our guide to assistive technology for dyslexia covers what is already on student devices.

Describe needs in general terms. “A student reading 2 grades below level” works just as well as a name and diagnosis, and keeps you within privacy rules.

Step 3: Fact-check everything

AI tools can state wrong facts with total confidence, invent quotes and cite sources that do not exist. Our explainer on AI hallucinations explains why.

A quick fact-checking routine:

1. Check facts against your adopted curriculum or a trusted reference. Dates, formulas, definitions and historical details are common error spots.
2. Work every math problem and answer key yourself. AI answer keys are a frequent source of mistakes.
3. Verify every source and link. If the AI names a book, article or website, confirm it exists and says what the AI claims.
4. Check standard alignment. Make sure the activity actually assesses the skill in the standard, not a nearby one.
5. Read for bias and cultural accuracy. Examples, names and scenarios should reflect your students and avoid stereotypes.

Key takeaway: AI drafts, you decide. Treat every AI lesson like a first draft from a capable but unreliable assistant: useful for speed, never ready for students without your review.

Step 4: Personalize and finalize

The last step turns a decent AI draft into your lesson.

• Add references to things your class cares about: a local event, a class joke, a recent field trip.
• Cut what will not work in your room, such as activities that need materials you lack.
• Adjust timing based on how your students actually work.
• Put the final version in your own template so it matches the rest of your materials.

This step usually takes a few minutes and makes the biggest difference in how the lesson lands.

Privacy and policy rules to follow

Student privacy is the most important guardrail.

Know FERPA basics. The Family Educational Rights and Privacy Act protects personally identifiable information in student education records. Sharing that information with an outside service is generally allowed only under specific conditions, such as when the vendor works under the school’s direct control through an agreement. This is not legal advice; your district’s guidance comes first.

Practical rules:

• Use district-approved tools whenever possible. They are more likely to be covered by data agreements.
• Never paste names, grades, IEP details or other identifiable student information into a consumer AI tool.
• Strip identifiers from student work before asking for feedback suggestions.
• Follow your district’s AI policy. The Gallup survey found that only 19% of teachers said their school had an AI policy, so ask your administration if you are unsure.

If your district is writing guidance, the structure in our AI acceptable use policy guide can be adapted for schools. The data-handling principles in is it safe to put company data into AI apply to student data too.

Where to start this week

1. Pick one lesson you teach next week that you would normally spend an hour preparing.
2. Use the Step 1 prompt template to draft it with an approved tool.
3. Ask for 2 differentiated versions of the main activity or reading.
4. Fact-check the draft and the answer key with the 5-point routine above.
5. Personalize it and teach it.
6. Note how long it took compared to your usual prep, and what you had to fix.

After 3 or 4 lessons, you will know which tasks AI handles well for your subject and which you are better off doing yourself.

Conclusion

AI will not plan your lessons for you, but it can do much of the drafting and differentiation that fills your evenings. Give it specific context, ask for multiple versions, check every fact and answer key, and finish with your own knowledge of your students. Keep student data out of unapproved tools. For a wider look at AI in schools, see AI in education, and for the other side of the classroom, read how teachers check for AI writing.

Sources

• Gallup: Teachers using AI tools save hours a week — 5.9 hours saved per week among weekly users; usage and school policy figures from the 2025 survey
• Khan Academy Blog: Khanmigo for Teachers now free for all US teachers — free access with Microsoft support and the teacher tools offered
• US Department of Education: FERPA — protections for personally identifiable information in education records