# 20 AI use cases for teachers

> Discover 20 AI use cases for teachers: lesson planning, activities, materials, feedback, assessment and student support.

- Site: Genialoh (https://genialoh.org)
- Language: en
- Category: Teaching with AI
- Reading time: 22 min
- HTML version: https://genialoh.org/#/en/blog/ai-use-cases-for-teachers

---
Artificial intelligence can help teachers plan lessons, create materials, generate examples, design activities and prepare feedback.

However, its usefulness does not depend solely on the quality of the tool.

It depends on the teacher defining a pedagogical goal, providing enough context, reviewing the outputs and keeping responsibility over content, assessment and academic decisions.

An AI-generated response can be:

- Clear, but incorrect.
- Well written, but shallow.
- Convincing, but based on invented data.
- Useful as a draft, but inappropriate for the group's level.
- Relevant in one discipline, but wrong in another.
- Creative, but incompatible with the academic program.

The best use cases are not those where AI replaces the teacher. They are those where it helps the teacher explore alternatives, reduce repetitive work and spend more time on mentoring, assessment and decision-making.

This guide presents 20 concrete applications that can be adapted to universities, high schools and preparatory programs.

## Before using AI: five basic questions

### 1. What academic goal am I pursuing?

The tool should support a learning outcome, an activity or a specific teaching need.

### 2. What information can I share?

Only the necessary, authorized context should be shared. Personal or sensitive student data should not be entered.

### 3. What do I expect to receive?

It can be an outline, a list of alternatives, a draft, a table or a set of questions. The clearer the expected format, the easier it will be to review the result.

### 4. How will I verify the answer?

The teacher needs to check concepts, sources, calculations, references, regulations or procedures.

### 5. What decision will I still make personally?

AI can propose. The teacher must select, correct, adapt and approve.

## 1. Create the initial structure of a class

AI can help organize a session by duration, topic, goals, student level, prior knowledge, modality, available resources and type of activity. It can suggest a sequence with prior-knowledge activation, problem framing, explanation, practice, discussion, formative assessment and closing.

> **Prompt example**
> 
> Propose a structure for a 90-minute university class on cost analysis. It targets second-semester students. The goal is that they distinguish fixed, variable and mixed costs and apply them to a simple case. Include an opening activity, explanation, team work and closing.

The teacher should review realistic timing, alignment with the goal, level fit, feasibility with available resources, student participation and whether the closing lets them observe learning. The output should be used as a draft, not a finished plan.

## 2. Formulate learning objectives

AI can transform general intentions into observable outcomes. It can turn «Students will know project management» into «Students will compare two project management methodologies and justify which one they would use in a given case».

> **Prompt example**
> 
> Rewrite these objectives using observable verbs. Students are undergraduates and the activity lasts two weeks. Avoid vague verbs like know, learn or understand.

AI can improve wording, but it does not decide what the student should learn.

## 3. Organize content by difficulty level

The teacher can use AI to prepare a progression from basic concepts to more complex applications: prior-knowledge recovery, initial understanding, direct application, analysis, integration, evaluation of alternatives and transfer to a new case.

> **Prompt example**
> 
> Organize these ten topics into a learning progression for first-semester students. Explain what prior knowledge each topic requires and propose a short activity to check the group is ready to advance.

## 4. Explain a concept in different ways

AI can generate different ways to explain the same topic: formal definition, introductory explanation, analogy, everyday example, professional case, step-by-step explanation, comparison and textual representation of a visualization.

> **Prompt example**
> 
> Explain the difference between correlation and causation in four ways: an academic definition, an explanation for a beginner, an example from education, and a counterexample.

A simple explanation should not eliminate essential nuances of the concept.

## 5. Generate examples and counterexamples

Examples allow applying a concept. Counterexamples help recognize its limits and detect errors. AI can propose correct examples, incorrect examples, ambiguous situations, professional applications, cases at different levels and frequent mistakes.

> **Prompt example**
> 
> Generate five examples and five counterexamples of conflict of interest for law students. Include academic, business and public-sector situations. Do not provide a conclusion; formulate a question to analyze each case.

## 6. Prepare case studies

AI can help create a first draft of a case for administration, law, engineering, psychology, education, communication, health sciences or architecture. The teacher can define fictional characters, restrictions, incomplete information and decisions to be made.

> **Prompt example**
> 
> Create a fictional case for business students about a medium-sized company that must decide between expanding its plant or outsourcing production. Include enough data to analyze costs, risks and capacity, but do not provide a solution.

> **Warning**
> 
> Real identifiable data should never be used without authorization.

## 7. Design simulations and role plays

AI can play a fictional counterpart: client, simulated patient, interviewer, user, negotiator, manager, committee member or case character. The student practices questions, decisions and communication.

> **Prompt example**
> 
> Act as a purchasing officer who has doubts about a supplier's proposal. Respond only with the case information and raise progressive objections. Do not immediately accept the student's recommendations.

Simulations in health, law or psychological intervention require extra precautions and do not replace supervised practice.

## 8. Create discussion questions

AI can propose questions of comprehension, application, analysis, comparison, argumentation, ethical evaluation, decision-making and professional reflection.

> **Prompt example**
> 
> Generate twelve discussion questions on the use of facial recognition. Order them from lowest to highest complexity and include technological, legal, ethical and social aspects.

## 9. Design learning activities

AI can generate ideas for case analyses, debates, projects, problem solving, source comparison, research, simulations, collaborative work, hands-on activities and reflections.

> **Prompt example**
> 
> Design a 50-minute activity for engineering students to compare three possible solutions to an energy-efficiency problem. They must use criteria, justify a recommendation and acknowledge limitations.

## 10. Adapt an activity to different levels

A teacher can prepare different versions of the same activity: introductory, intermediate, advanced, with more hints, with greater autonomy, with additional examples or with more complex restrictions.

> **Prompt example**
> 
> Create three versions of this problem: one with step-by-step support, one without hints, and a third with ambiguous information that forces students to state assumptions. Keep the same learning goal.

## 11. Prepare practice questions

AI can generate open-ended questions, multiple choice, true/false, problems, short cases, application questions and error-identification exercises.

> **Prompt example**
> 
> Generate ten multiple-choice questions on research methodology. Each question must have four options, one correct answer and an explanation. Avoid obviously wrong options.

Generated questions should not be used directly in high-stakes assessments without review.

## 12. Create rubric drafts

AI can help organize a rubric from goal, product, criteria, levels, weighting and evidence.

> **Prompt example**
> 
> Create a four-level rubric draft to evaluate an entrepreneurship project. Include problem definition, evidence, feasibility, risk analysis, communication and question response.

Words like «excellent» or «poor» need concrete descriptions to accompany them.

## 13. Review academic instructions

AI can serve as a second read to detect ambiguities, missing steps, contradictions, confusing dates, implicit requirements, overly technical language, unexplained criteria and assumptions.

> **Prompt example**
> 
> Review these instructions as if you were a first-semester student. Point out which parts could raise doubts, what information is missing and what questions a student would ask before starting.

## 14. Create supporting materials

AI can prepare drafts of study guides, glossaries, summaries, FAQs, review sheets, timelines, comparison tables, concept lists, organizers and introductions.

> **Prompt example**
> 
> Create a glossary of fifteen fundamental concepts for this unit. Each definition must be at most forty words and include a brief example related to the program.

## 15. Identify errors and misconceptions

The teacher can request an initial list of frequent errors related to concepts, procedures, calculations, data interpretation, source use, writing, code or argumentation.

> **Prompt example**
> 
> Identify ten common errors that beginner students may make when interpreting a hypothesis test. For each error, explain why it happens and propose a question that helps the student detect it.

The teacher may use anonymized examples but should never upload identifiable student work to unauthorized tools.

## 16. Prepare formative feedback

AI can help organize a draft comment with strengths, areas that need improvement, evidence, reflection questions, next steps and resources.

> **Prompt example**
> 
> Based on these criteria and this anonymized sample, propose a formative feedback draft. Separate strengths, areas to improve, one reflection question and one next step. Do not assign a grade.

Feedback should not be sent automatically. The teacher retains responsibility for reviewing, personalizing and approving each comment.

## 17. Redesign activities that can be fully solved by AI

Some traditional tasks can be almost entirely delegated. AI can help the teacher transform an activity to require research, verification, application, decisions, process evidence, reflection, oral defense, use of local cases and source comparison.

> **Prompt example**
> 
> This activity requests a general essay on leadership and can be easily solved by AI. Propose three ways to redesign it to observe the student's reasoning, preserve process evidence and allow declared use of AI.

## 18. Generate questions for an oral defense

The teacher can prepare questions to verify that the student understands the delivered work. They can explore decisions, sources, methods, limitations, assumptions, alternatives and application to a new situation.

> **Prompt example**
> 
> Generate eight questions for a five-minute oral defense of this project. Include one conceptual question, two on decisions, one on sources, one on limitations and three case variations.

An oral defense should verify comprehension, not become a punitive interrogation.

## 19. Analyze an AI-generated response as an academic activity

The AI response itself can become an object of study. Students can identify errors, verify claims, analyze biases, compare sources, detect omissions, improve the result and explain what they would accept or reject.

### Example activity

- Generate two responses on the same topic.
- Compare their arguments.
- Select five claims.
- Verify them.
- Identify errors.
- Create their own response.
- Declare the use of AI.
- Defend their decisions.

This kind of activity develops AI literacy, critical thinking, verification, responsibility and disciplinary judgment.

## 20. Create tutors or assistants contextualized by program

An institution can offer tutors organized around careers, grades, programs, areas, methodology, curricula and guidelines. The teacher can help define use cases, explanation level, questions, hints, limits, examples, situations requiring human help and verification messages.

Genialoh provides tutors configured by grade or career, AI support for teachers, personalized learning, reports and an experience aligned with each institution's educational project.

## Summary of the 20 use cases

| # | Use case | Initial output |
| --- | --- | --- |
| 1 | Structure a class | Session sequence |
| 2 | Formulate objectives | Observable objectives |
| 3 | Organize difficulty | Content progression |
| 4 | Explain in different ways | Alternative explanations |
| 5 | Create examples | Examples and counterexamples |
| 6 | Prepare cases | Case study |
| 7 | Design simulations | Interaction or role play |
| 8 | Create discussion questions | Question bank |
| 9 | Design activities | Activity draft |
| 10 | Adapt levels | Differentiated versions |
| 11 | Generate practice | Questions and exercises |
| 12 | Create rubrics | Criteria draft |
| 13 | Review instructions | Ambiguity detection |
| 14 | Prepare materials | Guides, glossaries and summaries |
| 15 | Anticipate errors | List of misconceptions |
| 16 | Draft feedback | Comment drafts |
| 17 | Redesign activities | Assessment with process evidence |
| 18 | Prepare oral defenses | Follow-up questions |
| 19 | Analyze AI answers | Critical activity |
| 20 | Create contextualized tutors | Program-based experience |

## How to write better prompts for AI

A useful prompt can include seven elements: role, context, goal, task, restrictions, format and criteria.

| Element | Example |
| --- | --- |
| Role | Act as an instructional design assistant. |
| Context | The activity is for third-semester business students. |
| Goal | They must compare alternatives and justify a decision. |
| Task | Propose three short cases. |
| Restrictions | Do not provide the solution and avoid personal data. |
| Format | Present each case in a table. |
| Criteria | Each case must allow more than one defensible answer. |

## What information a teacher must not share

- Full student names.
- Student IDs.
- Academic records.
- Individual grades.
- Medical information.
- Financial information.
- Identifiable personal situations.
- Family data.
- Passwords.
- Confidential exams.
- Unpublished question banks.
- Restricted research.
- Strategic institutional documents.
- Unauthorized materials.
- Company or patient information.

When it is necessary to use samples of student work, they must be anonymized. Anonymization is not just erasing the name; data that could indirectly identify the person must also be removed.

## What results the teacher must always verify

| Element | Required verification |
| --- | --- |
| Concepts and explanations | Check accuracy and depth |
| References | Confirm they exist and support the claim |
| Quotes | Compare with the original source |
| Data | Check origin, date and context |
| Calculations | Redo them or validate through an independent procedure |
| Code | Run it, test it and review its security |
| Regulations | Consult official applicable provisions |
| Professional cases | Confirm they represent practice adequately |
| Assessment materials | Review answers, distractors, difficulty and ambiguity |

## Uses that require special caution

- Automatic grading: AI should not make high-stakes academic decisions on its own.
- Feedback on identifiable work: identifiable student work should not be uploaded to unauthorized platforms.
- Psychological, medical or legal counseling: the tool does not replace professionals or institutional services.
- Fraud detection: an automatic score does not by itself prove an infraction.
- Performance prediction: usage patterns should not automatically become labels or decisions about students.
- Curricular recommendations: AI can propose alternatives, but program design belongs to academic leaders.

## Common teacher mistakes

- Copying the output without reviewing it.
- Using AI without a pedagogical goal.
- Sharing too much information.
- Requesting a full activity without criteria.
- Automating feedback.
- Creating assessments without checking the answers.
- Using the same case in any discipline.
- Confusing speed with quality.
- Not disclosing institutional use of AI.
- Expecting all teachers to adopt at the same pace.

## How to implement these cases institutionally

### 1. Select priority cases

It is not necessary to start with all twenty. The institution can choose three for planning, three for activities, two for assessment and two for student support.

### 2. Work with two programs

Selecting two careers, grades or areas allows testing disciplinary differences.

### 3. Create institutional examples

Examples must use the institution's programs, methodology, values, academic level and relevant cases.

### 4. Train teachers

Training must include practice, not only a general presentation.

### 5. Apply in one course

Each teacher can pick one use case and document goal, generated material, changes made, application, result and difficulties.

### 6. Share learnings

Coordinations can create a repository with useful prompts, reviewed materials, errors found, activities, rubrics and recommendations.

### 7. Evaluate and adjust

The institution can observe usefulness, preparation time, quality, participation, training needs, errors, incidents and cases worth expanding.

## Four-week teacher plan

| Week | Focus | Actions |
| --- | --- | --- |
| 1 | Explore | Select two cases, test prompts, identify errors, review privacy, pick one concrete application |
| 2 | Design | Prepare a material, review it, adapt it to the program, define rules, establish criteria |
| 3 | Apply | Use it in class, observe students, gather questions, document incidents, adjust |
| 4 | Evaluate | Review usefulness, compare to the previous method, gather feedback, share learnings, decide whether to continue |

## How to measure value for teachers

The university can observe trained teachers, applied use cases, materials created, redesigned activities, perceived preparation time, usefulness, quality, difficulties, errors found, support requests, satisfaction, continuity, cases shared among teachers and interested programs.

The number of prompts or conversations does not by itself demonstrate academic impact. It should be combined with material review, interviews, surveys and teacher observation.

## Frequently asked questions

### Can AI prepare a full class?

It can generate a draft structure, activities and materials. The teacher must review accuracy, level, timing and academic alignment.

### Can it create exams?

It can propose practice questions or drafts. High-stakes assessments require rigorous teacher review.

### Can it grade work?

It can help organize criteria or prepare initial comments, but assessment decisions must remain under human supervision.

### Can it generate feedback?

Yes, as a draft. The teacher must confirm it matches the work and that the tone is appropriate.

### Can teachers upload student work?

They should not upload identifiable work to unauthorized tools. Samples must be properly anonymized.

### Do teachers need to know how to code?

No. Most of these use cases can be done through natural-language prompts.

### Should all teachers use the same cases?

No. Each discipline, course and learning outcome needs different applications.

### How to avoid errors?

With context, clear criteria, verification, disciplinary review and an authorized platform.

### Does an institutional platform improve use cases?

It can keep activities and tutors aligned with the identity, methodology, values and programs provided by the institution.

### Does Genialoh include support for teachers?

Yes. The institutional proposal includes AI for teachers, tutors by career or grade, curricular contextualization, training, implementation, support and adoption follow-up.

### How can an institution get to know the platform?

It can start with a 30-minute meeting, select two careers or programs and provide authorized institutional documents to prepare a personalized demo.

## From individual use to an institutional teaching strategy

AI can help a single teacher prepare an activity or review instructions. But institutional value appears when use cases respond to the program, teachers are trained, authorized tools exist, information is protected, best practices are shared, there is support, adoption patterns are analyzed, results are verified and responsibility stays with people.

Genialoh enables a platform with institutional name, logo and colors; values and methodology; context from programs and curricula; tutors by career, grade or program; AI support for teachers; personalized learning; reports by student and group; institutional analytics; training; implementation; support and adoption follow-up.

> **Collaboration models**
> 
> The institution can choose between a tailored institutional license and a shared model with no direct institutional investment. In the shared model, students pay for access and the institution receives 10% of the corresponding revenue.

Book a first 30-minute meeting. During the conversation you will learn about Genialoh and can select two careers, grades or programs to visualize use cases adapted to your institution.

After receiving the full documentation, Genialoh can prepare a personalized functional demo in less than 14 days, depending on the agreed scope.