Teaching Students to Question AI: 12 Fact-Checking Activities That Build Critical Thinking

School Leadership 2.0 | Summary for Educators

Original Article:

Author: Med Kharbach, Ph.D. Source: Educators Technology Date: October 10, 2026

Topic Tags: Innovation & EdTech | Instruction & Pedagogy | Digital Literacy | Critical Thinking

THE BIG IDEA

Artificial intelligence can produce convincing answers in seconds. But how can students distinguish accurate information from fabricated facts, misleading statistics, or citations that don't actually support the claims being made?

In his article, Med Kharbach argues that fact-checking AI-generated information must become an essential part of digital literacy instruction.

He introduces a practical collection of 12 classroom activities designed for students in grades 6–12. These activities teach students to investigate AI-generated claims, verify sources, evaluate statistics, identify missing context, and revise inaccurate information.

The central message is that students should not simply accept or reject AI-generated answers. They should learn to investigate them systematically.

Importantly, these skills extend beyond artificial intelligence. The same habits help students evaluate social media posts, online news, research articles, and other digital information.

The objective is not to discourage AI use but to develop thoughtful, skeptical, evidence-driven learners who understand that a confidently written answer is not necessarily a correct answer.

KEY TAKEAWAYS

1. Teach a six-step fact-checking routine

Kharbach recommends a straightforward process that teachers can model across subjects:

2. Turn students into AI detectives

Several activities make verification engaging and manageable.

• AI Claim Audit: Students identify important claims in an AI response and determine which deserve investigation.

• Citation Detective: Students verify whether an AI-generated citation identifies a genuine publication.

• Real Source, Wrong Claim: Students examine whether a legitimate source actually supports what AI says it does.

• Trace the Statistic: Students investigate the origin, meaning, and context of numerical claims.

• Build a Better Answer: Students correct an unreliable AI response, replacing unsupported material with verified information.

These activities transform fact-checking from a vague instruction into an observable classroom skill.

3. Teach students that a real citation isn't enough

One of the most valuable lessons involves the distinction between a legitimate source and a legitimate claim.

An AI system might correctly cite a published research study while exaggerating its conclusions or misrepresenting what was investigated.

Students must learn to examine the original evidence, including the population studied, research methods, outcomes, and limitations.

4. Make uncertainty part of academic thinking

Students often assume that answers must be classified as either right or wrong.

Kharbach's approach introduces more sophisticated judgments. An answer may be partly correct, insufficiently supported, or misleading because important information is missing.

Learning to recognize uncertainty encourages intellectual humility and more careful reasoning.

5. Move beyond identifying errors to improving answers

The final goal is not simply to catch AI making mistakes.

Students should revise inaccurate responses, provide credible citations, and explain how their corrections improve the original answer.

This connects digital literacy directly to writing, research, and argumentation.

WHY IT MATTERS FOR SCHOOL LEADERS

Schools increasingly face a fundamental challenge: Students can generate polished explanations, essays, and research summaries without necessarily understanding or verifying the information they contain.

Banning AI does not solve this instructional problem.

Instead, schools need to establish clear expectations for responsible AI use. Verification should become a routine part of assignments involving AI-generated information.

The article also provides a valuable professional development opportunity. Teachers themselves can benefit from examining how AI citations, statistics, and research summaries sometimes misrepresent their sources.

These activities can be integrated into science, history, English, mathematics, and other subjects without creating an entirely new course.

LEADERSHIP ACTIONS

1. Establish a schoolwide AI verification routine. Encourage teachers to adopt common questions students must ask before relying on AI-generated information.

2. Demonstrate an AI fact-checking lesson at a faculty meeting. Present a convincing AI-generated paragraph containing several questionable claims and have teachers investigate them.

3. Build verification into assignments. Require students using AI for research to document how they checked key claims.

4. Integrate AI literacy across subjects. Science classes might examine research findings, history classes could verify quotations, and mathematics classes could investigate statistics.

5. Assess the verification process. Evaluate students not only on the accuracy of their final work but also on the quality of the evidence and reasoning used to reach their conclusions.

6. Provide ready-to-use resources. Share Kharbach's printable activity guide and encourage teachers to begin with one manageable lesson.

LEADER REFLECTION

Consider this question:

If a student submits a beautifully written, factually convincing AI-generated essay tomorrow, how confident are we that the student knows how to determine whether the information is actually true?

The challenge for schools is no longer simply teaching students how to find information. It is teaching them how to decide what information deserves their trust.

That may become one of the most important academic skills of the AI era.

Classroom Resource: Kharbach provides a downloadable guide with 12 activities, along with teacher notes, a source-quality checklist, a worked example, and student reflection materials.

Download the full guide

The full pack includes all 12 worksheets, along with a source quality checklist, a worked example, a student reflection page, and teacher notes. Other activities address quotations, images and captions, missing context, and comparisons between two AI answers.

The activities are organized into three levels: identify, verify, and evaluate and revise. You can choose a starting point that fits your students and curriculum. With limited device access, the guide suggests printing an AI response and relevant source excerpts so students can still complete the reasoning task.

Download the full AI Fact Checking Activities for Teachers and Students guide using the link in this post. Start with one activity, model the checking process, and give students time to explain what the evidence changed in their judgement.