# AI literacy for university students: what they need to learn

> Learn what university students need to know about artificial intelligence: fundamentals, verification, privacy, academic integrity, ethics and professional use.

- Site: Genialoh (https://genialoh.org)
- Language: en
- Category: AI literacy
- Reading time: 25 min
- HTML version: https://genialoh.org/#/en/blog/ai-literacy-university-students

---
AI literacy is becoming a cross-disciplinary competency for university students.

It does not mean everyone must learn to program models, study data engineering or become a technical specialist. It means students must understand how AI works, what it can be used for, what its limits are and what responsibilities remain with the person using it.

An AI-literate student should be able to:

- Use a tool with a clear purpose.
- Formulate appropriate instructions.
- Critically evaluate answers.
- Verify data, citations and references.
- Recognize errors and biases.
- Protect personal and institutional information.
- Disclose AI use in their work.
- Distinguish assistance from substitution.
- Apply the technology within their discipline.
- Explain and defend their own decisions.
- Recognize when to turn to a professor or specialist.

Universities should not limit themselves to allowing or banning tools. They must prepare students to use them with academic, professional and ethical judgment.

## What is AI literacy?

AI literacy is the set of knowledge, skills and attitudes that allow people to understand, use, evaluate and question AI systems. It includes four main dimensions.

| Dimension | Content |
| --- | --- |
| Understand | What AI and generative AI are, how they produce answers, why they can be wrong, the role of data, difference between search engine and generator, probabilistic answers. |
| Use | Define objective, provide context, formulate instructions, request formats, iterate, compare results, embed AI into a workflow. |
| Evaluate | Detect questionable claims, verify sources, review calculations, identify biases, recognize limits, question recommendations. |
| Act responsibly | Privacy, academic integrity, intellectual property, authorship, transparency, human responsibility, social and professional consequences. |

## Why should every student learn this?

AI is already present in academic and professional activities. Students find it in search engines, word processors, translation systems, design platforms, programming tools, data analysis, productivity software, customer service, recruitment, learning platforms, research and specialized tools of their profession.

AI literacy should therefore not be limited to engineering or computer science students. It is also relevant for business, law, medicine, psychology, education, communication, architecture, design, social sciences, humanities, finance and marketing. Each discipline requires its own applications, limits and criteria.

## What happens without AI literacy

- Excessive confidence in incorrect answers.
- Invented references.
- Copying content without understanding.
- Sharing sensitive information.
- Breaking academic rules.
- Tool dependency.
- Loss of fundamental skills.
- Poor professional decisions.
- Difficulty recognizing biases.
- Inability to explain submitted work.
- Use of unauthorized tools.
- Inequality between students with different levels of experience.

Literacy does not eliminate all these risks, but provides criteria to manage them.

## The 12 competencies students should develop

### 1. Understand the fundamentals of generative AI

Students need an accessible explanation of how answers are produced. They do not need to master the math of models, but they must understand that the tool identifies patterns, generates results from the given context, does not reason exactly like a person, does not automatically know the truth, can produce false information with apparent confidence, changes its answer depending on the instruction, may reproduce patterns present in its data and does not replace academic or professional sources.

> **Recommended activity**
> 
> Ask students to formulate the same question in three different ways and compare accuracy, depth, tone, assumptions and missing information. The goal is to show that how you ask affects the result, but does not guarantee it is correct.

### 2. Formulate purposeful instructions

Students should learn to include objective, context, audience, restrictions, format, criteria, allowed sources and level of depth.

**Instruction example**

| Version | Instruction |
| --- | --- |
| Basic | Explain inflation to me. |
| Contextualized | Explain inflation to a first-semester business student. Distinguish its main causes, use an example related to a Mexican company and point out two common mistakes when interpreting it. Present the answer in fewer than 500 words. |

They should learn to define the task, scope it, provide relevant information, avoid sharing unnecessary data, request alternatives, ask the tool to acknowledge limits and review the result before continuing.

### 3. Evaluate the quality of an answer

**Checklist**

| Criterion | Questions |
| --- | --- |
| Accuracy | Are the concepts correct? Are there questionable claims? Are there calculation errors? |
| Relevance | Does it answer the question? Does it include unnecessary information? Does it omit important elements? |
| Depth | Is it too superficial? Does it explain assumptions? Does it acknowledge exceptions? |
| Evidence | Does it provide verifiable sources? Do the references exist? Do they actually support the claims? |
| Context | Does the answer match the country, discipline and level? Does it use applicable regulations or data? |
| Responsibility | Does it communicate uncertainty? Does it recognize when professional intervention is required? |

Recommended activity: deliver an answer that mixes correct information, an invented fact, a nonexistent source, an unsupported conclusion and a useful example. Students must classify each element and explain how they would verify it.

### 4. Verify data, citations and references

Tools may generate nonexistent articles, incorrect authors, wrong dates, invented citations, outdated data, irrelevant links, non-applicable regulations and faulty calculations.

- Identify the important claims.
- Locate primary or trusted sources.
- Check the date.
- Compare the original content.
- Confirm authors, data and context.
- Remove what cannot be verified.
- Record the sources used.

**Verification table**

| Claim | Source consulted | Result | Decision |
| --- | --- | --- | --- |
| Claim 1 | Official document | Confirmed | Keep |
| Claim 2 | Academic article | Partial | Correct |
| Claim 3 | No evidence found | Not verifiable | Remove |

Students must cite the original source, not present an AI answer as a substitute for evidence.

### 5. Distinguish between a source and a support tool

| Role | Function |
| --- | --- |
| Tool to think with | Helps explore a topic. |
| Tool to produce with | Generates a draft or resource. |
| Source of evidence | Document, data or research that supports a claim. |

Universities must teach that a generated answer does not replace academic articles, books, databases, regulations, official reports, primary sources or verifiable data.

### 6. Protect personal and institutional data

> **Students should not share**
> 
> - Full names of others, student IDs, records, grades.
> - Medical or financial information.
> - Family data, passwords, identifiable photos without authorization.
> - Client data, patient information, confidential business documents.
> - Unpublished assessments, restricted research and unauthorized institutional materials.

Questions before sharing information: is it necessary?, can I anonymize it?, do I have authorization?, is the platform approved?, who could access it?, how long will it be kept?, what would happen if it were disclosed?

Recommended activity: present different scenarios and ask students to classify each as safe, needs anonymization, requires authorization or must not be shared.

### 7. Understand academic integrity

| Category | Meaning |
| --- | --- |
| AI not allowed | The activity must be completed without generative assistance. |
| Limited AI | It can be used in specific phases. |
| AI allowed with disclosure | Students may use it, but must disclose how. |
| AI required | Critical use of the tool is part of the learning. |

Students must understand that the same behavior can be acceptable in one activity and inappropriate in another. For example, generating an outline may be allowed in a project but forbidden in an individual assessment.

When submitting work, students must be able to explain it, defend it, identify sources, justify their decisions, recognize which parts were AI-supported, correct errors and apply the knowledge to another situation. Saying 'the AI generated it' does not remove their responsibility.

### 8. Disclose the use of AI

> **Short template**
> 
> I used [tool] for [purpose] during [stage]. I incorporated [elements]. I modified or rejected [explanation]. I verified the information through [sources or procedure]. I take responsibility for the final content.

Students should be able to explain what tool they used, for what, when, what instructions they gave, what they received, what they incorporated, what they corrected, how they verified and what limitations they found. The disclosure should be proportional.

### 9. Recognize biases and missing perspectives

Generated answers can reproduce stereotypes, generalizations, dominant perspectives, absence of local contexts, unequal representations, cultural assumptions, discriminatory language and non-applicable recommendations.

- Who is represented?
- Who is missing?
- What assumptions does it contain?
- Does the example match our context?
- Does the answer treat all groups fairly?
- What other perspectives should be consulted?
- What harm could a wrong application cause?

Recommended activity: request answers on the same case while modifying certain fictional variables and compare whether recommendations, language, confidence, assumptions and risk assessment change. Do not use real data or reinforce stereotypes during the activity.

### 10. Preserve own thinking and production

A tool can accelerate content generation, but universities must protect the development of fundamental skills.

**Gradual-support strategy**

| Step | Action |
| --- | --- |
| 1 | Try to solve it. |
| 2 | Identify the difficulty. |
| 3 | Ask for a hint. |
| 4 | Make a new attempt. |
| 5 | Request a partial explanation. |
| 6 | Verify. |
| 7 | Explain the procedure in your own words. |

The goal is not for AI to complete the task, but to help the student move forward.

### 11. Apply AI within a discipline

| Discipline | Typical applications |
| --- | --- |
| Business | Scenario analysis, proposals, market research, risk assessment, communication, verification of financial data. |
| Engineering | Explaining procedures, code review, simulation, error identification, assumption analysis, documentation. |
| Law | Organizing arguments, comparing concepts, reviewing texts, verifying regulations and case law, risk analysis. AI must never replace consulting current legal sources. |
| Health sciences | Simulated cases, communication, preparing questions, reviewing concepts, information analysis. Not a substitute for diagnosis, clinical supervision or professional care. |
| Education | Activity design, adapting explanations, rubrics, feedback, case analysis, digital literacy. |
| Communication and humanities | Comparing narratives, style analysis, structure review, identifying perspectives, preparing questions, creating and critiquing content. |

Each program must define which competencies matter for its profession.

### 12. Understand professional and social impact

Students must analyze how AI can change tasks, processes, jobs, responsibilities, services, evaluation of people, communication, decision-making, privacy and distribution of opportunities.

Using AI professionally implies reviewing results, documenting decisions, following rules, protecting information, recognizing limits, avoiding delegating critical decisions and being accountable for consequences.

> **Questions for professional reflection**
> 
> - Which tasks in my profession can be supported with AI?
> - Which must remain under human responsibility?
> - Which errors would have serious consequences?
> - What data should not be processed?
> - How would a decision be explained?
> - Who would be responsible in case of harm?
> - Which human capacities will be most important?

## Leveled curriculum proposal

| Level | Objective | Content | Product |
| --- | --- | --- | --- |
| 1. Fundamentals | Understand generative AI and use it in a safe, initial way. | Concepts, capabilities, limits, first instructions, verification, privacy, institutional rules. | Critical analysis of a generated answer. |
| 2. Academic use | Apply AI in study, research and production with transparency. | Search, organizing ideas, sources, disclosure, integrity, own thinking, data management. | Academic activity with logbook and verification. |
| 3. Disciplinary application | Use AI within the program with professional criteria. | Discipline cases, specialized tools, risks, validation, regulations, ethics, communication. | Project applied to a case in the program. |
| 4. Leadership and creation | Design solutions, evaluate tools and guide others. | Workflow design, platform evaluation, governance, tutor creation, impact analysis, risk management, responsible innovation. | Solution proposal or institutional project. |

## Eight-week basic course

| Week | Topic | Focus |
| --- | --- | --- |
| 1 | What is artificial intelligence | Concepts, types, general functioning, possibilities and limits. |
| 2 | Instructions and context | Objectives, structure, iteration, result comparison, criteria. |
| 3 | Verification | Data, references, citations, sources, calculations and evidence. |
| 4 | Privacy | Personal data, anonymization, sensitive information, approved platforms, risks. |
| 5 | Academic integrity | Use categories, disclosures, authorship, process evidence, consequences. |
| 6 | Biases and ethics | Perspectives, representation, responsibility, social impact, automated decisions. |
| 7 | Professional application | Cases per program, disciplinary risks, future work, human supervision. |
| 8 | Final project | Application, verification, disclosure, presentation and defense. |

## 30-day intensive plan

| Days | Focus |
| --- | --- |
| 1–5 | Fundamentals: what AI is, how it generates answers, possibilities, errors, first practice. |
| 6–10 | Critical use: instructions, comparison, verification, sources, quality. |
| 11–15 | Responsibility: privacy, integrity, disclosure, authorship, biases. |
| 16–20 | Academic application: study, research, writing, programming, activities. |
| 21–25 | Disciplinary application: cases per program, risks, professional criteria, approved tools. |
| 26–30 | Evaluation: project, logbook, verification, presentation, defense. |

## Ten activities to develop AI literacy

- Detect errors: analyze an incorrect answer and correct it.
- Verify references: check which sources exist and which were invented.
- Compare instructions: observe how the answer changes with context.
- Analyze biases: identify assumptions and missing perspectives.
- Create a disclosure: document how the tool was used.
- Improve an answer: turn a superficial result into a verified explanation.
- Solve with hints: use AI as a tutor without asking for the final answer.
- Compare generic and institutional tool: analyze differences in context, methodology and usefulness.
- Evaluate a professional case: determine when AI can help and when human supervision is required.
- Build a portfolio: gather instructions, results, errors, verifications and reflections.

## How to assess the competencies

Assessment should not focus on memorizing definitions. It may consider understanding, formulating instructions, verification, source quality, data protection, bias identification, transparency, own production, disciplinary application and ability to explain.

**AI literacy rubric**

| Criterion | Initial | Developing | Competent | Advanced |
| --- | --- | --- | --- | --- |
| Understanding | Uses the tool without recognizing limits | Recognizes some risks | Explains capabilities and limits | Analyzes implications and guides others |
| Instructions | Formulates general requests | Adds basic context | Defines objective, format and criteria | Designs iterative processes |
| Verification | Accepts results | Checks some data | Verifies claims and sources | Designs a rigorous procedure |
| Privacy | Shares information without evaluation | Recognizes sensitive data | Minimizes and anonymizes | Assesses complex risks |
| Integrity | Unaware of rules | Discloses partially | Applies rules transparently | Helps design responsible practices |
| Critical thinking | Accepts the recommendation | Identifies basic problems | Compares, questions and justifies | Assesses professional consequences |
| Application | Uses general cases | Adapts examples | Applies disciplinary criteria | Designs responsible solutions |

## Literacy portfolio

- First generated answer.
- Corrections.
- Sources consulted.
- Verification table.
- Use disclosures.
- Discipline cases.
- Bias analysis.
- Reflections.
- Final project.
- Evidence of defense.

The portfolio makes it possible to observe the development of judgment over time.

## How to implement literacy institutionally

| Step | Action |
| --- | --- |
| 1 | Define common competencies in a cross-disciplinary framework. |
| 2 | Adapt it per program: applications, risks and evidence. |
| 3 | Train faculty to teach and assess these competencies. |
| 4 | Establish approved tools. |
| 5 | Update rules: each activity states the allowed use level. |
| 6 | Create materials: guides, examples, cases, disclosures, rubrics and FAQs. |
| 7 | Start with two programs to compare disciplines and adjust. |
| 8 | Measure results: trained students, verification, disclosures, incidents, activities, perception. |
| 9 | Update periodically: tools, practices and risks change. |

## The role of an institutional platform

An institutional platform can help so that literacy does not depend on isolated tools. It can include university identity, values, methodology, programs, tutors per program or grade, common guidelines, use cases, faculty support, reports, analytics, training and support.

Genialoh is configured with the identity, vision and methodology of each school or university, and includes tutors related to the educational offering, faculty support and personalized learning. The proposal organizes tutors by grade for high school and by program for universities, using the curriculum, programs, values and methodology provided by the institution.

This can help students practice within a common academic context, while all answers still need review and verification.

## Indicators to measure the program

| Category | Indicators |
| --- | --- |
| Participation | Enrolled students, completion, participation per program, submitted activities. |
| Competencies | Verification quality, correct sources, data protection, complete disclosures, error identification, disciplinary application. |
| Behavior | Use of approved tools, questions about rules, incidents, need for coaching, ability to explain work. |
| Perception | Critical confidence, usefulness, clarity of rules, professional preparation, intent to keep learning. |
| Institution | Trained faculty, participating programs, activities created, tutors configured, documented cases, updated rules. |

The goal must not be to reach the highest number of AI conversations. It must be to develop students able to use AI with judgment.

## Frequent mistakes when teaching AI literacy

- Teaching only how to write prompts.
- Presenting the tool as infallible.
- Focusing only on prohibitions.
- Allowing everything without conditions.
- Using the same examples across all programs.
- Not teaching source verification.
- Not working with real cases.
- Assessing only through an exam.
- Not training faculty.
- Ignoring privacy.
- Confusing frequent use with competence.

## Frequently asked questions

### Should every university student learn about AI?

Everyone should develop basic literacy. The depth and applications must be adapted to each program.

### Do they need to learn programming?

Not necessarily. Basic literacy focuses on understanding, use, evaluation and responsibility. Some programs will require additional technical training.

### What should they learn first?

That a generated answer can be useful and convincing, but not necessarily correct. They must learn to verify it.

### How is academic integrity taught?

Through visible rules, use categories, disclosures, examples and activities that preserve evidence of the process.

### Should students submit all their conversations?

Not always. Evidence must be proportional. Main instruction, incorporated result and process explanation can be requested.

### How to avoid dependency?

Through non-AI activities, gradual support, prior attempts, oral defense and reasoning assessment.

### How is literacy assessed?

With hands-on activities, source verification, error analysis, disclosures, disciplinary projects and ability to explain decisions.

### Can it be taught as a course?

Yes. It can also be integrated across several subjects. A mix of general training and per-program application is usually more effective.

### What information must they not share?

Personal data, records, grades, medical or financial information, passwords, restricted research and unauthorized documents.

### Does an institutional platform replace training?

No. It provides a common environment, but students still need teaching, practice, rules and supervision.

### How can a university start?

Define basic competencies, train faculty, select two programs and evaluate a first institutional experience.

## Training critical users, not just frequent users

AI literacy is not about students producing more content in less time. It is about being able to understand, ask, verify, compare, protect, disclose, decide, apply, explain and take responsibility.

Genialoh lets you create an institutional platform with name, logo and colors; values and methodology; context based on plans and programs; tutors per program, grade or level; AI support for faculty; personalized learning; reports per student and group; institutional analytics; training; implementation; support; and adoption tracking.

The university can start with two programs to design literacy experiences adapted to its context.

> **Book a first meeting**
> 
> Genialoh can prepare a functional demo with the institutional identity and the two selected programs in less than 14 days after receiving complete documentation. The timeline depends on the agreed scope. After the presentation, up to five authorized people can explore the demo for 30 calendar days.