AI can reduce the time it takes to move from an idea to an organized first draft. It cannot know whether that draft fits your students, schedule, curriculum, school policy, or professional judgment unless those constraints are made explicit—and the final result is reviewed by an educator.
This workflow treats AI as a drafting and revision partner. It keeps the teacher responsible for the instructional purpose and the final classroom-ready plan.
1. Define the learning result before the activity
Start with what students should know, understand, or be able to do by the end of the lesson. A useful input names the grade or course, the learning target, relevant standard or curriculum expectation, and the level of thinking students should demonstrate.
Too open-ended
“Create a lesson about fractions.”
More useful
“Create a 45-minute grade 5 lesson in which students compare fractions with unlike denominators and justify their reasoning with visual models.”
A precise outcome helps the assistant select activities for a reason. It also gives the teacher a clear standard for evaluating the draft.
2. Supply the constraints of the real classroom
Include the information that would change the plan if it changed. Depending on the lesson, that may include:
- Available minutes and where the lesson falls in the unit
- Class size, grouping options, and available materials or technology
- What students already know and common misconceptions
- Required standards, vocabulary, texts, or district templates
- Documented supports that may be described without exposing unnecessary personal information
- What students must produce before the lesson ends
Use the minimum student information needed for the task. Follow school policy and avoid entering identifiable student information into any tool that has not been approved for that use.
3. Ask for a lesson architecture, not a wall of text
Request a structure you can inspect quickly: opening or retrieval, brief modeling, guided practice, independent or collaborative practice, checks for understanding, closure, and an extension or next step. Ask for estimated time and required materials beside each segment.
For a first draft, it is often better to request concise teacher moves, student actions, and decision points than a word-for-word script. The result stays editable and leaves room for professional voice.
4. Plan evidence of learning throughout the lesson
A polished activity sequence is not enough. Ask what evidence will show whether students are making progress toward the target.
- Add one or two questions that reveal the most likely misconception.
- State what a strong response should include.
- Identify what the teacher should do if the check shows widespread confusion.
- End with a short product, explanation, performance, or exit task tied directly to the objective.
This turns the generated plan into a decision-support draft rather than a schedule of activities.
5. Differentiate deliberately
A generic request to “differentiate” often produces a generic list. Name the barrier or learner need and ask for a change that preserves the learning target.
Access
Vocabulary preview, visual models, chunked directions, worked examples, read-aloud support, or multiple ways to engage with the content.
Expression and challenge
Sentence frames, oral or visual response options, strategic grouping, reduced cognitive clutter, extension questions, or transfer to a new context.
Review every suggestion against the learner, the goal, and any required plan or accommodation. AI output is a draft, not a determination about a student.
6. Run a teacher review before using the plan
Read the plan once for instructional coherence and once for classroom reality.
- Alignment: Do the activities and final evidence match the stated learning goal?
- Accuracy: Are the content, examples, answers, and vocabulary correct?
- Timing: Can this group realistically complete the sequence?
- Access: Are supports specific, respectful, and connected to the barrier?
- Assessment: Will the checks reveal understanding rather than only completion?
- Safety and policy: Are materials, data use, and student-facing AI interactions appropriate and authorized?
- Teacher voice: Does the plan sound and feel like something you would actually teach?
Revise the weakest section rather than regenerating everything. Targeted follow-ups—“shorten the opening to five minutes,” “replace the final task with an oral explanation rubric,” or “add one misconception check”—usually preserve more of what already works.
A reusable AI lesson-planning prompt
Adapt this template to your subject, grade, and planning format. Remove any field that does not apply.
Act as a lesson-planning partner. Create an editable first draft for educator review. Course or grade: Learning target: Standard or curriculum expectation: Lesson length and place in the unit: What students already know: Likely misconception: Available materials and technology: Access needs or constraints (without unnecessary personal data): Required student product or evidence: Organize the plan into: opening/retrieval, modeling, guided practice, independent or collaborative practice, checks for understanding, closure, and next step. For each segment, include the estimated time, teacher move, student action, and evidence to look for. Flag any assumption you had to make. Do not invent student data, citations, standards, or school policy.
Apply the workflow in Faraday AI
Jasper can help educators draft lessons, objectives, materials, checks for understanding, accommodations, and follow-up artifacts inside a focused teaching workspace. Learn more about the Faraday AI teaching assistant, or open the public guest tour.