ADL Logo
ADL Logo
Lesson Plan

Becoming a Critical Thinker in an AI-Driven World: Media Literacy Strategies for the Classroom 

Teacher in classroom pointing to screen of laptop students are working on

ADDITIONAL TERMS FOR EDUCATORS:

Generative Artificial Intelligence (GAI): Unlike more traditional AI systems, which are typically designed to recognize patterns in existing data and make predictions based on those patterns, generative artificial intelligence (GAI) systems are designed to produce new content in response to user prompts. These outputs can include text, images, audio or video content.

Examples of popular GAI chatbots include Open AI’s ChatGPT and Anthropic’s Claude. Popular GAI video generators include Google’s Veo 3 and OpenAI’s Sora 2. 

Large Language Models (LLMs): Are AI models which have been trained on massive amounts of data through machine learning to process and generate human-like text, allowing them to answer complex queries and perform tasks like text generation, summarization and translation.

LLMs also power some of the most commonly used chatbots. For example, GPT-5 and GPT-4.1 are two of several LLMs that have powered ChatGPT. 

As Artificial Intelligence (AI) continues to dominate conversations in the media and across industries, educators have an opportunity to bring the conversation into their classrooms. The surge of AI affords educators a necessary and urgent opportunity to strengthen critical thinking skills, promote and demonstrate responsible AI use and to emphasize that AI is a tool—not the final say—that can be utilized to support and augment learning. Explore the content below for media literacy lessons designed to bring conversations about AI and its intersection with antisemitism and bias into classrooms.

Background Information

  • Definition: Artificial Intelligence (better known as AI) refers to computer systems that can perform tasks that typically require human intelligence, like answering questions, generating text, creating images, or making recommendations. 
  • The Data: AI is already being used by young people. In a 2025 study, RAND found that 54% of students and 53% of teachers in core subjects reported using AI for school, and College Board found that 84% of high school students reported using generative AI for schoolwork. 
  • Video: Common Sense Education--What is AI? (2:37), Oxford University--How Does AI Learn? (1:35) or Kurzgesagt--A.I. ‐ Humanity's Final Invention? (16:42) 
    • Ensure students understand how AI works: Data enters in → model learns patterns → decisions/ outputs. 
      • **Suggestion for teachers: consider beginning the conversation by asking students: What does AI mean to you? (You will likely have students reference popular AI tools like ChatGPT, Claude, Google’s Gemini, etc. Consider drawing student attention to AI features that have been built into social media platforms as well.) 

  • Antisemitism refers to prejudice, discrimination, or hostility directed at Jewish people. It is generally based on negative stereotypes, myths, or misinformation about Jews, Judaism, or Jewish identity. It manifests in harmful beliefs, attitudes, language, exclusion, harassment, and violence against individuals or Jewish communities.  
  • A note for educators—if students ask why focus on antisemitism and AI: Antisemitism shares characteristics with racism, anti-Muslim bias and other forms of hate, but it also has unique features that make it harder to identify, especially in AI-generated content. Unlike most forms of bias, antisemitism portrays Jews as simultaneously too powerful and too weak, as both a threat and a target. Antisemitism morphs to fit the values and fears of any given context, which means it can camouflage itself in ways that generic bias detection tools often miss. 

  • AI and LLMs (Large Language Models) recognize and run off of patterns: Many AI and LLMs are created by scraping the internet and looking for repeated information or patterns. Given the prevalence of hate, bias and antisemitism online, it is no wonder that these then show up in AI-generated content and responses. 
  • AI chatbots can repeat harmful stereotypes: AI chatbots learn from huge amounts of text written by humans. If that text includes antisemitic content, the AI can absorb those ideas and repeat them. For example, a chatbot might generate responses that include false, harmful stereotypes about Jewish people, even without being directly asked to. 
  • AI image and video generators can create hateful content: Tools that generate images or videos can be used to create antisemitic cartoons, memes or fake videos (called "deepfakes") that mock, dehumanize or spread lies about Jewish people. These images can go viral and reach millions of people. 
  • AI can create “rabbit holes:” A lot of platforms like social media use AI to recommend content to keep people engaged. If someone watches or clicks on one piece of hateful content, the algorithm might keep recommending more and more extreme material, pulling that person deeper into antisemitic ideas without them even realizing it. 
  • AI can speed up hate: People who want to spread antisemitism can use AI tools to quickly produce large amounts of hateful text, fake news articles, memes or social media posts. AI makes it faster and easier to flood the internet with this harmful content. 
  • AI search tools can elevate harmful websites: Some AI-powered search engines pull answers from websites that may include antisemitic conspiracy theories. If AI presents false information as a straight-up "answer," people doing research might believe it without questioning the source. 

  • It hides in plain sight. Antisemitic ideas often use coded words or familiar-sounding arguments instead of obvious slurs. AI is generally better at catching explicit hate speech than subtle bias. 
  • It borrows from history and politics. When antisemitic tropes are dressed up as historical facts or political commentary, they can sound reasonable — and AI often cannot tell the difference. 
  • AI learns from the internet — bias and all. AI tools are trained on enormous amounts of online content, and that content already contains a lot of antisemitism. If the training material is biased, the AI can be too. 
  • Context matters — and AI struggles with context. Whether something is antisemitic often depends on how it is framed, who is saying it and why. That kind of nuanced judgment is hard for AI to make. 
  • Jewish identity is complex. Antisemitism can target Jewish people as a religion, a culture or an ethnicity — sometimes all three at once. AI tools that do not understand that complexity are more likely to miss it. 

The bottom line for students: Just because AI did not flag something as harmful does not mean it is accurate or fair. While AI responses may sound authoritative, and they may give the appearance of knowledge, the reality is LLMs are tools. And a tool is only as good as its user. That is exactly why human judgment—and media literacy—matter more than ever. 

For more information, explore ADL’s AI Index, Artificial Intelligence: What do Parents, Caregivers and Educators Need to Know? or take ADL Education’s course for educators: Teaching with AI. 

Lesson: Building Media Literacy Skills—Tracing the Claim 

A quick note: The content of this lesson has been created for students who already have an understanding of what AI is, how it “learns” and how bias can enter. Consider using the Background Information above to supplement student learning. For learning for educators, consider ADL’s course Teaching with AI: Antisemitism, Jewish Identity and the Holocaust (with professional development credits). 

Students will be able to... 

  • Learn how to use lateral reading as a strategy to evaluate AI-generated content. 
  • Reflect upon how they can use lateral reading (and other media literacy skills) to be more informed and critical users of AI-generated content. 

Introduction: Begin by asking students: have you ever seen a headline or statistic come up on social media or in an answer from an AI chatbot, and later found out it wasn’t true? How did you learn that and how did it change your thinking about the topic? 

  • Allow for 2-3 responses before transitioning to the next part of the lesson. 

Trace the Claim: Explain to students what it means to trace a claim generated by AI. 

  • Suggested script: When encountering a striking statistic or headline, it is important to trace it back to its original source rather than accepting it at face value. Claims are frequently repeated across multiple websites until their origins become unclear or distorted. Ask yourself: who first made this claim, and what evidence did they provide? If a clear and credible original source cannot be identified, the claim should be treated with significant skepticism. 

Model for Students: 

  • Distribute to each student a copy of the example passage: A Viral Statistic About Nobel Prizes and Jewish Achievement. Explain to students that you are going to model for them how you would evaluate an AI-generated paragraph to trace the claims made in it. Direct students to take notes when you do on their own copies of the passage. 

  • If technology allows, project your screen—you will want the passage up and an Internet tab. Read the passage aloud (using the Teacher Guide to help) and pause at different claims to search for their credibility and reliability. Add your own annotations to your shared passage for students to see—pointing out the places where AI is deliberately misleading and shaping claims to create a specific narrative and false sense of credibility. 

  • Transition by asking students: how did your view of the passage change when we traced the claims in it? Why is it important to verify sources and information in AI-generated content? 

Small Group Practice:  

  • Break students into small groups of approximately 4-5, and give each a passage (repeat passages as needed). Allow students approximately 10 minutes to work in their groups to trace the claims in their assigned passages. 

  • Allow each group to share the key takeaways that they found in their passages, asking each group how their view of the passage changed after they verified the claims in them. 

Closing Discussion: Close by asking students to reflect and debrief on their learning from the activity. Some questions might include: 

  • Why is it important that we verify the claims and information presented by AI? 

  • Did any of the false claims surprise you? Did you learn anything from researching the claims? 

  • Why do you think misinformation so often uses the names of real institutions, real journals or real publications alongside fabricated details? 

  • Why might AI be particularly likely to produce content that sounds authoritative but contains false sources? 

  • What steps can you take to ensure you are a more critical and informed user of AI moving forward? 

Lesson: Debating Policy: Hot Topics 

A quick note: The content of this lesson has been created for students who already have an understanding of what AI is, how it “learns” and how bias can enter. Consider using the Background Information above to supplement student learning. For learning for educators, consider ADL’s course Teaching with AI: Antisemitism, Jewish Identity and the Holocaust (with professional development credits). 

Students will be able to... 

  • Reflect and discuss challenging thought topics related to AI, its regulation and its potential to generate biased content.

Introduction: Explain to students that through this lesson, they will discuss and consider “hot topics” that come up with regard to regulating AI and its potential to generate biased content.

  • Allow for 2-3 responses before transitioning to the next part of the lesson. 

Pre-Reading: Based on the topic of the conversation, provide students with an article or two to read ahead of the conversation. Some suggested articles are: 

Discussion Protocol: Use the discussion protocol of your choice. You might consider a Socratic Seminar (NCTE guidance), fishbowl (Edutopia guidance), or other protocol depending on your students and classroom procedures. Based on the protocol, select one to all of the questions below: 

  • Should AI be regulated by international government laws, independent ethical boards, religious/moral guidelines or the free market? 
  • When an AI system generates harmful content, who is morally responsible: the engineers who built the system, the users who prompted it, those who generate content on the web or the company that released it? 
  • What are some ways that consumers and the public can push tech companies to act ethically? (I.e. public pressure, laws and policies, etc.) 
  • We’ve discussed a lot of the negative aspects of AI. What are some positive ones? How can we ensure that we are ethical and critical users of AI moving forward? 

Elementary Guidance

As AI becomes an increasing part of the world we live in, elementary educators may want to discuss AI and how bias can potentially enter. It is recommended that the focus more is on how AI functions and how it has the potential to produce negative outcomes. Consider using a metaphor like that of a toolbox: 

  • Think of online spaces and programs as a toolbox. AI is just one tool in the toolbox. Sometimes, it will be helpful—like when you need a flathead screwdriver to build furniture. And sometimes, it isn’t the right tool—thinking about the different screwdrivers you might need to build a piece of furniture. And sometimes, it can be too much and cause harm—like if you swing a hammer too hard and hit your finger or the wall. 
  • AI isn’t something to be feared or run away from, but you should learn how to use it so you can make sure you are using the right tool at the right time. 
  • Consider framing the conversation by sharing an age-appropriate video like Sesame Workshop’s AI and Us.