SStretchLearn
Sign inMembershipStart learning
Catalog / Tech & AI / AI Tools for Teachers & Educators
Tech & AIBeginnerPreview

AI Tools for Teachers & Educators

A practical, classroom-tested course that takes K-12 and post-secondary educators from curiosity to a real AI workflow: lesson plans aligned to standards, rubrics that survive a department review, materials differentiated for five reading levels, feedback that is faster but still yours, and the admin work that eats your evenings. Every lesson uses named tools and ends with a copy-ready prompt, a real example, and the safety guardrails that keep student data and academic integrity intact.

For K-12 and post-secondary teachers, instructional coaches, special educators, and school leaders who want a real AI workflow for planning, assessment, differentiation, feedback, and admin, with no coding required.

Course content

The AI Toolkit Built for Teachers45m
Protecting Student Data: FERPA and the Privacy Rules50m
Prompting Basics: Role, Context, Task, Format45m
Generating Standards-Aligned Lesson Plans50m
Hooks, Activities, and Differentiation with UDL45m
Unit Planning, Pacing, and Backward Design50m
Creating and Calibrating Rubrics50m
Writing Assessments with Bloom's and Webb's DOK50m
Differentiating One Text for Every Reader50m

Workbook & downloads

Put the course into practice — a printable workbook plus editable templates you can fill in and reuse.

Download workbook (PDF)19 KBDownload (XLSX)9 KBDownload (XLSX)7 KBDownload (DOCX)8 KB
Preview the workbook
This workbook turns the course into a working classroom AI practice: a vetted tool-and-privacy plan, a reusable prompt bank for planning and feedback, calibrated rubrics and DOK-mapped assessments, a differentiation workflow, and parent-communication and admin templates that stay inside policy. Work one section per module, doing each build step in a real tool with de-identified material and filling in every worksheet. By the end you will have artifacts you can use the next morning, not just notes, and a habit of using AI hard while protecting student data and your professional judgment.

Getting Started Safely: Tools, Privacy, and Prompts

Choose one general assistant and one teacher-built tool your district allows, lock in your FERPA de-identification habit, and master the Role-Context-Task-Format prompt before you automate anything.
Worksheet: Tool Selection and Approval Sheet
Decide the two tools you will actually standardize on and confirm both are cleared by your district before you build any workflow on them. Depth in two approved tools beats a shallow tour of a dozen apps, and the approval column is the one that keeps you out of trouble.
  • General assistant I will use (ChatGPT / Claude / Gemini)
  • Teacher-built tool I will use (MagicSchool / Diffit / Brisk / SchoolAI / Khanmigo)
  • District-approved? (yes / no / need to ask) for each tool
  • Data agreement in place for student-data use? (yes / no)
  • Primary jobs each tool will do (planning / rubrics / differentiation / feedback / admin)
Exercise: De-Identify a Real Prompt
Take a request you would genuinely make to an AI and strip every detail that could identify a student or your school, then compare the two versions. This builds the muscle that keeps you FERPA-safe under time pressure. Do this with a real example you would otherwise have typed unsafely.
  1. Write the unsafe version of a prompt as you might first think it, then circle every identifying detail (name, school, class, ID, disability).
  2. Rewrite it fully de-identified, replacing names with Student A and describing needs generically (reads two grades below level).
  3. Confirm the de-identified version still gives the AI enough context to be useful.
  4. Decide which approved tool you would use if the task genuinely required real, identifying details.
Exercise: Practice the Role-Context-Task-Format Prompt
Convert three vague requests into structured prompts using all four parts, run them, and note the quality jump. Keep the strong versions for your prompt bank. The extra forty seconds of typing is what saves twenty minutes of fixing.
  1. Rewrite make a lesson on fractions as a full Role, Context, Task, Format prompt with your grade, period length, and class reality.
  2. Rewrite make a quiz on a topic you teach, specifying the Bloom's mix, number of questions, and an answer key.
  3. Rewrite write a worksheet, specifying the format (table, one page) and a model style to match.
  4. For one prompt, run it, then refine the output with a single follow-up (easier reading level, add an example) instead of restarting.
Checklist: Privacy Gut-Check Before You Hit Enter
  • Every real name, student ID, and identifying detail has been removed
  • The tool is district-approved for the type of information being entered
  • I would be comfortable if this exact text appeared in a public records request
  • The actual grade, placement, or official note stays in my hands, not the AI's
  • I could honestly explain to a parent how their child's data is handled

Planning Lessons and Standards-Aligned Materials

Generate standards-aligned lessons and units the AI can draft fast, design for engagement and UDL on purpose, and use backward design so every day builds toward the assessment.
Worksheet: Lesson Plan Input Sheet
Before prompting for any lesson, fill in the inputs you would give a student teacher so the draft arrives close to teachable. Paste the standard verbatim, not just a topic. Reuse this sheet every time you generate a plan.
  • Exact standard or objective (full text, e.g. NGSS MS-LS1-6, Common Core, state/provincial)
  • Grade level and period length in minutes
  • Class reality (size, range of reading levels, supports needed)
  • Lesson structure to use (5E / gradual release / workshop)
  • Non-negotiables the plan must include (hook, hands-on, exit ticket, vocabulary)
Exercise: Build a UDL Option Menu for One Objective
Pick one learning objective and use AI to generate engagement, representation, and expression options, then choose the ones you will actually use. Designing the three multiples in is far easier at the draft stage than bolting them on later.
  1. Give me five different hooks for this objective, including a question, a short demo, and a real-world story.
  2. Present this concept three ways: a plain-language text, a described visual or diagram, and an analogy a student this age would relate to.
  3. Give me three ways students can demonstrate this learning: written, drawn, and recorded or built.
  4. Generate sentence starters, a simple-definition vocabulary list, and a graphic organizer for students who need scaffolding.
Worksheet: Backward-Design Unit Planner
Plan a unit in the correct order: results first, evidence second, daily lessons last, then fit it to your real calendar. Hold the end goal fixed and let the AI propose the sequence. Flag where the pacing is tight so you can add buffer.
  • Stage 1 — Standards and 2-3 enduring understandings / essential questions
  • Stage 2 — Summative assessment or performance task that proves those understandings
  • Number of class days available (and known interruptions)
  • Stage 3 — Day-by-day lesson sequence (each day's focus)
  • Alignment note — which part of the final task each day prepares students for
Checklist: Professional-Filter Checklist for Any AI Plan
  • The lesson actually teaches the named standard and does not drift to an easier skill
  • Every content claim, date, and fact has been verified before teaching
  • Pacing is padded for real transitions and students, not the AI's optimistic timing
  • Materials, examples, and contexts have been swapped for ones my students and room have
  • Each lesson in a unit genuinely builds toward the summative task

Assessment, Rubrics, and Differentiated Materials

Draft rubrics in seconds and calibrate them on real samples, write assessments that span Bloom's and Webb's DOK levels, and turn one text into leveled, translated, supported versions for every reader.
Worksheet: Rubric Specification Sheet
Specify the rubric you actually want so the AI drafts a usable grid, then you edit for your expectations rather than building from nothing. Demand observable descriptors, not vague words like good or poor, at every level.
  • Assignment and its goal (e.g. 4-paragraph persuasive essay)
  • Criteria that matter most (trimmed to 3-4)
  • Scale (number of levels and their names, e.g. Exceeding / Meeting / Approaching / Beginning)
  • Standard to align the top level to (full text)
  • Observable descriptor for the Meeting level on each criterion
Exercise: Calibrate the Rubric on Anonymized Samples
Test your draft rubric against two or three de-identified student samples so scoring is consistent and defensible before you grade a stack. Every place your score and the AI's disagree is a vague descriptor to tighten. Use only de-identified work.
  1. Apply this rubric to this anonymized student sample and explain your score for each criterion.
  2. Identify which descriptors are too vague to score consistently (watch for words like good, clear, strong).
  3. Rewrite each fuzzy descriptor to name what a reader would actually see (e.g. at least two pieces of evidence, each explained).
  4. Re-score the same sample with the tightened rubric and confirm your score and the AI's now match.
Exercise: Raise an Assessment's Cognitive Level
Take a quiz that is all recall and deliberately spread it across thinking levels using Bloom's or Webb's DOK, then verify the labels and the key. Naming the framework in the prompt is what lifts an assessment off the recall floor.
  1. Revise this quiz so the questions span DOK 1 recall, DOK 2 applying, and at least two DOK 3 strategic-thinking items, and label each.
  2. Make sure wrong multiple-choice options reflect real, common student misconceptions, not obvious throwaways.
  3. Provide an answer key with the reasoning for each correct answer.
  4. Check that each question's labeled level is honest and that every item maps to something I actually taught.
Worksheet: One-Text Differentiation Plan
Plan how a single source text will reach every reader in the room using Diffit, Brisk, or a leveling prompt. Read each generated version to protect accuracy and rigor, and distribute discreetly so no student is singled out by which version they receive.
  • Source text or URL and its original reading level
  • Reading levels to generate (e.g. 3rd, 5th, original) and which students get each
  • Home-language supports needed (translated summary, bilingual glossary) and for whom
  • Comprehension supports to attach (vocabulary, leveled questions, sentence starters)
  • Accuracy and rigor check — confirmed simpler versions kept key ideas and challenging questions

Feedback, Communication, and Administrative Tasks

Draft specific, growth-oriented feedback that you still own, produce calm parent messages and accessible IEP-friendly supports while protecting privacy, and hand the recurring admin tail to AI to win hours back.
Worksheet: Feedback Drafting Brief
Set up your feedback prompt so drafts are specific, warm, and actionable, and so you keep the grade and the judgment. Only de-identified work goes into a general tool, and you read and personalize every comment before a student sees it.
  • Rubric or criteria to ground the feedback
  • Feedback shape (e.g. 2 specific strengths, then 2 concrete next steps)
  • Tone and reading level for the student audience
  • Specificity rule (point to actual moments in the work, name an actionable next step)
  • My ownership step (who assigns the grade, what I verify and personalize)
Exercise: Draft a Sensitive Parent Email
Use AI to produce a calm, professional first draft of a message home about a concern, then adjust and send. A good first draft makes a hard conversation easier. Keep it de-identified in a general tool, or use an approved tool if real details are required.
  1. Draft an email to a parent about a missing-work pattern that leads with a positive, states the issue plainly, and ends with a clear next step.
  2. Rewrite it to be warm and non-accusatory while keeping it honest.
  3. Produce a version in the family's home language for me to spot-check with a fluent speaker.
  4. Turn my conference notes into talking points organized as strengths, growth, and a plan.
Exercise: Make a Task IEP-Friendly
Use AI to create accessible supports that match an accommodation, while remembering the IEP team, not the AI, decides the accommodations. Keep the content de-identified, and finalize any identified version in your official systems.
  1. Rewrite this multi-step direction as short, numbered steps with one action each.
  2. Generate sentence starters, a word bank, and a model response for a student who needs written-expression support.
  3. Produce a reduced-item version of this assignment that keeps the same learning target with fewer questions.
  4. Add clear headings, plain language, and described diagrams to make this material easier to comprehend.
Checklist: Admin Time-Win Checklist
  • Picked one recurring task I dislike (newsletter, sub plans) to let AI draft for a month
  • Saved the prompts that worked into a small, reusable prompt library
  • Used AI to summarize at least one long policy or thread into key action items
  • Drafted a sub plan with timing, fallback activities, and a feedback note
  • Skimmed, fixed, and personalized anything going out under my name before sending

Your Action Plan

  1. Choose one general assistant and one teacher-built tool, and confirm both are district-approved before building any workflow.
  2. Lock in the de-identification habit: rewrite one real prompt to strip every identifying detail and keep it as your model.
  3. Adopt Role-Context-Task-Format as your default prompt and bank the strongest versions for reuse.
  4. Fill the Lesson Plan Input Sheet and generate one standards-aligned lesson you can teach the next morning.
  5. Build a UDL option menu for one objective and choose the engagement, representation, and expression paths you will use.
  6. Plan one unit with backward design: results, then the summative task, then the day-by-day sequence fit to your calendar.
  7. Draft a rubric, then calibrate it on two or three anonymized samples and tighten every vague descriptor.
  8. Revise one quiz to span Bloom's or Webb's DOK levels, label each item, and verify the labels and the key.
  9. Differentiate one source text into leveled and translated versions, reading each to protect accuracy and rigor.
  10. Set up your feedback brief and admin prompt library, then let AI draft feedback, parent messages, and one weekly admin task that you own and personalize.

Pairs well with

Courses members commonly take alongside this one.

Flagship CoursePreview

Freelance Business Foundations: Position, Price, Sell, and Deliver High-Value Services

Freelancing · Beginner · 16h

Build a freelance business clients understand, trust, and pay for—without vague positioning, random referrals, or underpriced custom work.

Self-pacedPreview
Client GrowthPreview

Freelance Client Acquisition: Outreach, Leads, Referrals, and Deal Flow

Freelancing · Beginner · 15h 30m

Build a repeatable acquisition system that turns targeting, outreach, referrals, and follow-up into a stable freelance opportunity pipeline.

Self-pacedPreview
Sales SystemPreview

Freelance Sales & Proposals: Discovery Calls, Scoping, Objections, and Closing

Freelancing · Intermediate · 16h

Run better discovery calls, scope work properly, write proposals clients can decide on, and close without discounting your value into the floor.

Self-pacedPreview