From ChatGPT to Critical Thinking: How UAE Schools Can Build Real AI Literacy with Mexty

From ChatGPT to Critical Thinking: How UAE Schools Can Build Real AI Literacy with Mexty

 

AI is already part of school life.

Students use it to search, write, translate, summarise, generate images and solve problems. It is tempting to call that AI Literacy but It is not.

Knowing how to ask ChatGPT a question is useful, but real AI Literacy begins when a student can ask a different set of questions:

Why did the model give me this answer?
What evidence supports it?
Could it be biased?
What might be missing?
Should I trust it?

That shift from simply using AI to questioning, evaluating and governing it is becoming central to the UAE approach to AI education. The UAE framework is designed as a structured K–12 progression, moving students from foundational awareness toward critical evaluation, responsible use, design and real-world application, with human judgment and oversight remaining important throughout.

The challenge for schools is no longer only to introduce AI.

It is to turn that ambition into a coherent learning journey that teachers can actually deliver, adapt and assess.

That is where Mexty can support implementation.

 

AI Literacy is not the same as AI usage

A school can have coding classes, robotics clubs and access to generative AI and still not have a coherent AI Literacy program.

Students may become confident AI users without necessarily understanding how AI works, why outputs can be unreliable, how bias can appear, when information should be verified or when human judgment needs to override an AI recommendation.

Clearly isolated activities may be useful, but they do not automatically create progression. A stronger model develops AI Literacy progressively, from simple ideas about machines and patterns in the early years to data, algorithms, bias, misinformation, hallucinations, privacy, accountability and governance in later grades.

The learning journey becomes:

Discover → Understand → Use → Question → Evaluate → Create → Govern

Mexty helps turn that progression into an actual learning structure.

 

Start with ready-made UAE AI Literacy learning paths

Schools should not need to build every AI Literacy lesson from scratch.

Mexty provides editable AI Literacy learning paths from KG to Grade 12. Each grade can already include the relevant:

Learning Path → Courses → Interactive Activities → AI Literacy Lab Experiences → Assessments

The school therefore starts with a structured foundation rather than an empty authoring screen.

A Grade 8 learning path, for example, may already include courses and activities around:

• generative AI;
• misinformation;
• bias;
• deepfakes;
• privacy;
• data quality;
• trusted information;
• human oversight.

The role of the teacher is not to recreate the whole sequence.

It is to review, adapt and enrich it.

 

Adapt the learning path to the school

The UAE learning paths provide the structure, but schools still need flexibility.

A school could open the Grade 7 UAE AI Literacy template and:

• edit existing content or add its own text, images, videos and resources;
• adapt the sequencing to fit the school timetable or curriculum structure;
• replace or enrich examples with cases relevant to the school community;
• adjust the difficulty or depth of courses and activities;
• add the school’s own digital citizenship or AI-use policies;
• add or modify interactive activities and AI Literacy Lab exercises;
• customise the final assessment.

This is important because UAE schools operate across different curricula, school models and pedagogical approaches.

Mexty provides the structured foundation.

The school keeps pedagogical control.

 

Turn explanations into interactive experiences

AI Literacy is especially difficult to teach through passive content alone.

Students can read a definition of hallucination without being able to recognise one.

They can memorise the meaning of bias without being able to identify bias in an AI output.

They can learn that sources should be verified without knowing how to verify them.

Mexty allows teachers to turn these concepts into interactive learning experiences.

For example, instead of simply explaining hallucinations, a teacher could use an activity where students receive an AI-generated answer containing several plausible claims and are asked:

Which statement needs verification?

A bias lesson could become a categorisation exercise:

Potentially biased
 Needs more evidence
 Acceptable with justification

A privacy lesson could become a branching scenario:

A student wants to paste personal information into an AI tool. What should they do next?

Interactive activities make students practise judgment rather than simply memorise terminology.

 

Use the AI Literacy Lab to let students investigate AI

One of the strongest ways to teach AI Literacy is to let students observe AI behaviour directly.

Mexty’s AI Literacy Lab is designed for this purpose.

Students can run controlled experiments such as:

Compare different AI models

Ask the same question to two different models.

Then examine:

• what information each model included;
• what each model omitted;
• how confident the answers appear;
• whether the conclusions differ;
• which answer is better supported by evidence.

The objective is not to decide that one model is always better.

It is to understand that different AI systems can behave differently.

Change the context

Ask: Should schools allow students to use AI for homework?

Then ask the same question with different contexts: “Answer as a teacher”  and “Answer as a student”. You could also add: Answer as a school principal.

Students can then compare how the different perspectives affect:

• the benefits highlighted;
• the risks mentioned;
• assumptions about responsibility;
• concerns about fairness or cheating;
• the importance given to learning, efficiency or control.

The learning objective is to show that context and perspective can shape an AI-generated response, even when the underlying question stays the same.

 

Investigate hallucinations

Students can be given a question where the model may produce an unsupported claim.

The learning task is not simply to find the error.

It is to ask:

Why does the answer sound convincing?
 What should be verified?
 What evidence would we need?

Compare general AI with trusted-source-grounded AI

Students can ask a general model a question and then compare the response with one grounded in approved school materials.

This helps them understand the difference between:

general model knowledge and trusted-source-grounded knowledge.

 

Ground learning in trusted school knowledge

A central AI Literacy skill is understanding that a fluent answer is not necessarily a reliable answer.

Mexty Knowledge Bases allow schools to bring approved content into the learning environment.

Teachers can add:

• school policies;
• curriculum documents;
• approved articles;
• digital-safety guidance;
• teacher-created resources;
• reference documents.

These can then be used as trusted sources for courses and activities.

A teacher could, for example, upload the school’s own AI-use policy and create an activity asking students to evaluate different AI-use scenarios against that policy.

Or a teacher could provide an authoritative source and ask learners to compare it with a general AI-generated explanation.

This makes source verification practical.

 

Add teacher-created content without rebuilding the course

The ready-made learning paths are not intended to lock teachers into fixed content.

Teachers can add their own:

• PowerPoints;
• PDFs;
• videos;
• classroom instructions;
• examples;
• discussion prompts;
• worksheets;
• assessment questions.

Existing school material can therefore become part of the AI Literacy journey rather than sitting outside the platform.

A teacher might already have a PDF called: Responsible Use of AI at Our School

That document could be used to enrich the relevant course, create an interactive activity or become part of an assessment.

This helps schools combine their existing resources with the structured AI Literacy learning path.

 

Teach students to question AI, not simply prompt it

Prompting is useful but prompt engineering alone is not AI Literacy.

A student may know how to produce an impressive answer while having no idea whether that answer is accurate, biased or appropriately sourced.

This is why the Mexty approach places greater emphasis on questioning AI.

A student should learn to ask:

• Why did the model produce this answer?
• What information influenced it?
• What evidence supports it?
• What might be missing?
• Could another model answer differently?
• Could the response be biased?
• Does the answer need human review?

The goal is to move students from using AI toward understanding when it should be questioned and when human judgment must come first.

 

Assess judgment, not only recall

If AI Literacy is treated as a real curriculum, schools need evidence that students are progressing.

A multiple-choice quiz may tell you whether a learner knows what a hallucination is.

It does not necessarily tell you whether the learner can recognise one.

Mexty Evaluations can combine different assessment formats.

For example:

Knowledge check: What is an AI hallucination?

Visual recognition: Show a response and ask students to identify the unsupported claim.

Categorisation: Students classify responses as:

Students classify each AI-generated response according to whether it appears trustworthy, requires further verification, or is unsafe to use without human review.

Decision scenario: A model recommends an important action.

The student must decide whether to accept the AI recommendation, verify it further, or escalate it to a human for review, and then indicate how confident they are in that decision.

This adds an important metacognitive dimension.

The assessment should move beyond simple multiple-choice questions toward decision scenarios, visual recognition, categorisation, confidence checks and comparison activities.

 

Build teacher readiness alongside student learning

A successful AI Literacy program depends on teachers.

Teachers do not need to become AI engineers but they do need enough confidence to guide students through questions such as:

• Why can AI hallucinate?
• How does context influence a response?
• What makes a source trustworthy?
• How can bias enter an AI system?
• What information should students never share?
• When is AI assistance appropriate?
• When must a human remain responsible?

Mexty can support teacher development through dedicated learning paths.

Teachers can also experience the same AI Literacy Lab activities before using them with students.

That gives teachers both content knowledge and practical familiarity with the learning experience.

[Explore AI Literacy Teacher Training →

 

Keep governance inside the learning experience

AI governance should not exist only as an IT policy because students need to understand why rules exist. A school should define:

• which AI tools students can use;
• which models are appropriate for different ages;
• what student information can be shared;
• which sources are trusted;
• how AI-generated content is reviewed;
• when teachers must intervene;
• how academic integrity is handled.

Mexty helps schools integrate these principles into the learning environment rather than treating governance as a separate document.

The school can add its own AI-use policy to a course.

Students can work through scenarios based on that policy.

Teachers can use trusted sources.

Assessments can test whether learners understand when human review is required.

This turns governance into part of AI Literacy itself.

A practical Grade 9 example

Imagine a Grade 9 learning path focused on:

LLMs, Prompting and Trust

The template already contains the course and related activities.

A teacher might adapt it as follows.

1. Review the existing lesson

The teacher checks the explanations of LLMs, hallucinations and prompting.

2. Add the school AI policy

The school’s rules around responsible AI use and academic integrity are added as trusted content.

3. Adapt one example

A generic prompt is replaced with a classroom-relevant example.

4. Run an AI Literacy Lab

Students ask:

Is social media beneficial for teenagers?

They compare two models.

They then change the context.

5. Add a verification task

Students must identify which claims require external evidence.

6. Run the assessment

The evaluation combines a short quiz, a comparison task and a decision scenario.

7. Review student evidence

The teacher can see whether students understand the concepts and whether they can apply judgment in practice.

That is a complete AI Literacy learning experience. Not just an AI demonstration.

 

From AI usage to AI capability

The UAE direction is not simply about giving students access to more AI tools. It is about building a progressive capability.

Young learners begin with awareness.

Older learners start understanding patterns and data. Then they question outputs, compare models, examine evidence, evaluate bias and understand governance.

Discover → Understand → Use → Question → Evaluate → Create → Govern

Mexty helps schools turn that progression into something operational:

Ready-made KG–12 Learning Paths
 Editable Courses
 Interactive Activities
 AI Literacy Labs
 Trusted Knowledge
 Assessments
 Teacher Training
 Learner Evidence

The objective is not to replace the UAE framework, teachers or school leadership. It is to provide the learning infrastructure that helps schools move from policy and curriculum objectives to classroom practice.

The most important outcome is not whether students can use AI faster. It is whether they learn when to use it, when to question it, when to verify it and when human judgment must come first.