Civic Education in the Age of AI: Should We Trust AI-Generated Lesson Plans? 

Every teacher plans lessons for students, but the way they go about doing it can vary significantly. Some veteran educators, drawing on years of teaching experience, plan out classes in their heads, while many new-to-the-profession educators spend long hours researching and planning each instructional activity. Some rely on district-purchased prepackaged curriculum materials to guide their plans while others turn to the Internet to find resources, activities, and assessments. Many strive to weave premade curriculum materials, their own teaching experiences, and ideas from the Internet into each day’s class experiences.  

Now, seemingly overnight, generative artificial intelligence (GenAI) technologies have introduced a new dimension to lesson planning for students (Extance, 2023). With a simple prompt (e.g., “create a lesson plan about the Declaration of Independence”), GenAI chatbots can produce, in a matter of seconds, a full-class lesson plan that includes activities, assessments, content for teacher-led presentations, discussion prompts, and homework ideas. 

Numerous teachers are already turning to GenAI tools for lesson planning, raising the following questions: What kinds of lesson plans, instructional activities, and learning assessments are AI chatbots generating for civic education? And should we trust AI-generated lesson plans? 

Ideally, civics lesson plans should engage students in higher order thinking activities to prepare them for their future roles as informed decision-makers in democratic settings. Civics learning experiences should also incorporate diverse, multicultural, and inclusive content to ensure students understand and consider all people’s voices and experiences in ongoing discussions and debates over public and educational policies.  

For this study, we examined 310 AI-generated lesson plans, featuring a total of 2,230 activities for civic education. The lesson plans were designed based on the 53 standards in the Massachusetts eighth-grade United States and Massachusetts Government and Civic Life content standards (Massachusetts Department of Education, 2018). The Massachusetts PreK-12 History and Social Science standards, including these eighth-grade standards, received an exemplary rating from Thomas B. Fordham Institute — one of only five state standards nationally to be so recognized (Stern et al., 2021).  

Our focus on one state’s curriculum standards is based on an increasing emphasis on standards-based curriculum in schools across the nation. Throughout our state, teachers are expected to show administrators that their lesson plans address state standards. The Massachusetts Department of Elementary and Secondary Education (2024), has stated that it expects educators to use state learning standards “to inform their instruction” (para. 2). Nationally, 39 states have laws requiring schools to align curriculum with content standards, while 11 more recommend alignment (Ballotpedia, 2024). A recent survey found that teachers reported increased use of standards-aligned curriculum materials over previous years (Turna, et al., 2022).  

Given the increased role of curriculum standards in the delivery of educational learning for students, we chose to examine the design, quality, and content of AI-generated lesson plans that are aligned to state standards. The following research questions guided our study:  

• RQ1: What type of student thinking skills are promoted in AI-generated lesson plans for the Massachusetts eighth-grade United States and Massachusetts Government and Civic Life content standards? 

• RQ2: In what ways do AI-generated lesson plans for the Massachusetts eighth-grade United States and Massachusetts Government and Civic Life content standards integrate multicultural, diverse, and inclusive perspectives? 

Review of Literature 

The professional work of civics, history, and social studies teachers is filled with complex and competing demands. In addition to being strapped for time and trying to balance work and personal lives, these educators must try to incorporate the latest historical research, ever-changing current events, and new policy proposals and laws into their lessons. All of these conditions often happen alongside changes in curriculum frameworks and shifting educational requirements. As a result, many teachers turn to the Internet for help with lesson planning. Teachers have reported getting premade lesson plans, prepackaged teaching materials, ideas for learning activities, and other curriculum resources from a variety of digital sources (e.g., Pinterest, Teachers Give Teachers); but the most common source that teachers have turned to over the past decade is Teachers Pay Teachers (TPT), an online marketplace where teachers can buy and sell educational materials. 

While high school history teachers also draw curriculum resources from highly regarded sites such as the Smithsonian Institution, PBS Learning Media, National Geographic, Learning for Justice, and the Zinn Education project, one recent study found more than half of history teachers used materials from TPT (American History Association, 2024). Examining self-reports from more than 1,300 educators, researchers noted many teachers go to TPT to find instructional resources when their districts do not provide them or they do not have the time to find them on their own (Carpenter & Shelton, 2022).  

Teachers often adapt premade instructional resources from TPT, social media, or the Internet, rather than using them as is; however, one study found that teachers tend to make adaptations focused on students’ needs, level of engagement, and abilities rather than improving the quality or diversity of the content (Schroeder et al., 2019). Several scholars have examined the instructional quality, content, and standards-alignment of TPT resources and noted that these materials are often of low to moderate quality (Harris et al., 2023; Polikoff & Dean, 2019), lack alignment with standards (Aguilar et al., 2022), and can be harmful to student learners (Harris et al., 2023; Shelton et al., 2020). One study found TPT instructional materials related to Martin Luther King, Jr., Black History Month, and the Civil Rights Movement had problematic historical narratives, lacked rigor, and did little to “help students understand the significance of Dr. King’s work or the movement for civil and human rights” (Rodriguez et al., 2020, p. 510).  

For civic and government learning, TPT lesson plans and related curriculum materials tend to create highly monocultural and in-group views of citizenship by emphasizing patriotic American symbols and topics (i.e., U.S. flag, Liberty Bell, Statue of Liberty) while ignoring the past and present-day experiences of historically marginalized and disenfranchised groups (Stebbins et al., 2024). Looking at TPT alongside other online marketplaces like TES and Amazon Ignite, researchers declared, “Teachers beware and vet with care” (Archambault et al., 2021; title).  

AI for Lesson Planning   

Whereas lesson planning with TPT and other online lesson-sharing sites involves searching for or browsing premade lessons created by other educators, AI lesson planning involves teachers prompting a GenAI tool to generate a lesson plan from scratch based on the chatbot’s training data. GenAI technologies are dramatically different from traditional AI tools in that they are designed to generate new content, including text and media, rather than provide predictions or recommendations or perform a specific task.  

ChatGPT was the first publicly available GenAI technology. Several other GenAI tools have come out since, including Google’s Gemini, Microsoft’s Copilot, Meta AI, Claude, Perplexity, and Pi.ai. There are also educator-specific GenAI tools like Magic School, Eduaide.ai, Brisk, and Diffit that purport to save teachers time by generating lesson plans, assessments, instructional materials, and other resources. One developer of an education-focused GenAI tool recently asserted that “in the future teachers won’t have to write lesson plans,” and AI tools will help teachers “spend less time behind your desk and more time in front of students” (Khanmigo, personal communication, March 13, 2025). 

The number of teachers using GenAI rose steadily throughout 2023 and 2024 (Will, 2025). One recent study found nearly half of all teachers reported using GenAI tools at least weekly (Rosenbaum, 2024). Interestingly, while teachers have received little, if any, training on how to use GenAI tools, they still report using these tools in the classroom with the most common uses including (a) exploring new ideas for teaching (41%); (b) creating lesson plans (38%); (c) creating material to present to students during class (36%), and (d) creating student assignments (35%; Langreo, 2024). 

As increasingly more teachers turn to GenAI for lesson planning, resources, and instructional support, questions about the quality and reliability of AI-generated materials have moved front and center in educational research. GenAI systems can generate information that is fake, false, and flat out wrong — what has been called the “next misinformation nightmare” (Gold & Fischer, 2023). Additionally, GenAI chatbots are not good at history, offering incorrect facts and failing to provide in-depth information about the causes and consequences of events. Scholars have also found that chatbots perform worse for certain regions of the world, notably Sub-Saharan Africa (Rollet, 2025).  

GenAI tools have been known to produce harmful, stereotypical, and biased outputs based on class, race, gender, language, and dis/ablity (Larsen-Scheuer, 2023). Bias can appear in many forms: through harmful images; unfair and incomplete histories and experiences of Black, Latino/Latina, Indigenous, and Asian peoples; and broad and inappropriate gender stereotypes; and discriminatory divides between standard and nonstandard speakers of English, as well as between dominant and less dominant languages in global society (Lazaro, 2023; Rettberg, 2024; Ta & Lee, 2023). These multiple concerns have motivated our investigation into the kinds of civic learning experiences GenAI platforms are providing to teachers and students. 

Studies of GenAI as a lesson planning tool are only beginning to be published. One study found that preservice teachers used GenAI tools for personal and academic tasks, but less commonly for lesson planning (Hsu et al., 2024). While Moran and Coward (2024) found that using GenAI as a lesson plan designer enhanced the self-confidence of preservice English teachers preparing learning experiences for students, they also found that the teacher candidates in the study expressed reservations about chatbot errors and concluded that more detailed prompts were needed to generate more classroom-ready lessons. Moundridou et al. (2024) noted that educators can use GenAI tools to design and manage each phase of the inquiry-based learning process. Meanwhile, Clark and van Kessel (2024) wrote that, while AI-generated lessons might introduce teachers to new ideas and possible source materials, the “uncritical use of AI carries with it the danger of slipping into problematic discourses” (para. 77). Collectively, these studies demonstrate that more investigations are needed to understand the types of teaching and learning experiences GenAI tools are producing. 

Guiding Theoretical Frameworks 

To address our research questions, we used two guiding theoretical frameworks to examine AI-generated civics and government lesson plans: (a) Bloom’s Taxonomy (Bloom et al., 1956)  to assess the level of thinking promoted in each plan and (b) Banks’ (1999) Four Levels of Integration of Multicultural Content to examine the incorporation of diverse, multicultural, and inclusive content and perspectives in the plans.  

Bloom’s Taxonomy 

Bloom’s Taxonomy is a widely known framework organized around student thinking skills. It was first introduced in 1956 and subsequently revised in 2001 (Armstrong, 2010). The framework distinguishes between different ways of thinking ranging from “lower order” thinking skills (remembering, understanding, applying) to “higher order” thinking skills (analyzing, evaluating and creating). Students utilize different thinking skills and engage in different actions at each level in the revised Bloom’s framework, as follows: 

• Remember: Recall facts and basic concepts. 

• Understand: Explain ideas or concepts.   

• Apply: Use information in new situations. 

• Analyze: Draw connections among ideas. 

• Evaluate: Justify a stand or decision. 

• Create: Produce new or original work. 

In the context of civic learning, asking students to remember names, dates, and places emphasizes lower order thinking by focusing students’ attention on recalling information. By contrast, higher order thinking happens when students learn by evaluating information and generating new and original materials.  

Bloom’s Taxonomy has been used in multiple disciplines to evaluate and revise learning outcomes by examining the cognitive demands of learning activities and aligning them with educational goals (Barari, et.al., 2020; Karanja & Malone, 2021; Soleimani & Kheiri, 2016; Su & Osisek, 2011). For example, Sultana and Klecker (1999) found that 1^st-year teachers often design lessons featuring learning objectives at the lowest cognitive level. Researchers have also used the taxonomy to analyze patterns in evaluation data — such as recurring learning activities at different levels of thinking skills — to not only help pinpoint gaps in curriculum design but also ensure learning activities progress toward critical thinking.  

Banks’ Four Levels of Integration of Multicultural Content Model 

Banks’ (1999) Four Levels of Integration of Multicultural Content Model measures teachers’ use of multicultural history content along a four-part scale ranging from contributions at its lowest level through additive, transformation, and then social action at the highest level. A contributions approach basically is confined to acknowledging the contribution of particular individuals or groups of individuals, for example incorporating Martin Luther King Jr. into a conversation about civil rights activists or discussing the roles of women suffragists during Women’s History Month. Focusing on heroes and holidays, special days, and a few change makers, a contributions approach does not engage in broader investigations of continuity and change in history. 

The additive approach features an increased focus on diverse histories throughout the curriculum and across the school year but does not alter the dominant narratives found in most textbooks and curriculum frameworks. In an additive approach, as James Loewen (2018) has pointed out, issues in history tend to appear at certain points in the curriculum and then disappear as if the problem has been resolved. For example, the African American Civil Rights Movement is discussed when learning about the 1950s and 1960s, but students do not learn about the long struggle for civil rights in the past or the continuing challenges to equality and social justice today. While commonly used, both contributions and additive approaches are a largely insufficient means of recognizing diverse histories.  

The transformation approach includes featuring the histories of underrepresented and excluded groups as those peoples experienced those histories (Hickman & Porfilio, 2012). A transformative study of westward expansion, for example, would include not only accounts of settlers, wagon trains, the Gold Rush and the transcontinental railroad, but sources and stories of indigenous peoples whose lands were taken, Chinese railroad workers who faced discrimination, formerly enslaved African Americans who found free land and new lives on the frontier, and women who played multiple roles in native and immigrant communities. Bringing the voices, experiences, and histories of marginalized individuals and groups directly into classroom learning enables students to discover how diverse people experienced the past.  

The social action approach includes creating opportunities for students to address historical wrongs and current-day social problems through civically engaged actions. Social action is a core idea in civics education, where it is often called action civics. Action civics refers to students becoming actively engaged in learning about and responding to social and community problems while advancing goals of racial, gender, and socioeconomic equity and justice for everyone. Civic learning through social action, stated the National Council for the Social Studies, enables students to “not only to study how others participate, but also to practice participating and taking informed action themselves” (quoted in Singer, 2019, para. 8).  

Banks’ multicultural model has been a longstanding framework for examining the integration of diversity, equity, and inclusion topics in academic curriculum and student learning experiences in multiple fields and classroom settings (Bagnardi et al., 2009; Brown & Livstromb, 2020; Henson, 2015; Lu et al., 2022). Banks (1995) recognized that the “academic achievement of students of color and low-income students can be increased when teaching strategies and activities build upon the cultural and linguistic strengths of students, and when teachers have cultural competency in the cultures of their students” (p. 393). Examining AI-generated lesson plans using Banks’s model can provide a framework for assessing how clearly those plans integrate diversity, equity, and inclusion topics and multicultural perspectives in student learning experiences. 

Research Methods 

This qualitative, exploratory study was designed to evaluate the quality of AI-generated lesson plans for the Massachusetts eighth-grade United States and Massachusetts Government and Civic Life content standards. For the study, AI-generated lesson plans served as data that were examined and interpreted based on the previously mentioned theoretical frameworks. The purpose of this study was to provide educators with evidence to inform their use of GenAI tools for lesson planning.  

Data Collection 

To collect data for this study, we prompted three GenAI chatbots — ChatGPT, Gemini, and Copilot — to generate lesson plans for each of the eighth-grade Government and Civic Life standards in the Massachusetts History & Social Science Curriculum Framework (Massachusetts Department of Education, 2018). This curriculum has 53 standards distributed across seven major topics:  

• Topic 1: The Philosophical Foundations of the United States Political System (5 standards);  
• Topic 2: The Development of the United States Government (5 standards);  
• Topic 3: The Institutions of United States Government (5 standards);  
• Topic 4: Rights and Responsibilities of Citizens (13 standards);  
• Topic 5: The Constitution, Amendments, Supreme Court Decisions (6 standards with the sixth standard having 3 additional sub standards);  
• Topic 6: The Structure of Massachusetts State and Local Government (10 standards);  
• Topic 7: Freedom of the Press and News/Media Literacy (6 standards). 

We selected ChatGPT, Gemini, and Copilot to generate lesson plans for this curriculum based on the popularity and perceived capabilities of these tools. ChatGPT, created by OpenAI, is a GenAI chatbot designed to produce text (and now visuals) in response to a prompt from a user. For this study, we used the GPT4o model of ChatGPT released in May 2024. This model has been widely recognized for its conversational abilities, adaptability, and structured responses.  

Gemini, Google’s version of a large language model, similarly produces text and visuals in response to a prompt from a user. We used the Gemini 1.5 Flash model released in September 2024. This model was designed to focus on contextual relevance and efficiency, providing a rapid response style that aligns with instructional prompt-based tasks.  

Copilot is Microsoft’s version of a GenAI chatbot, which can produce text and visuals. Unlike the other platforms, Copilot does not have a specific version/model. As it explained for itself, “I am continuously updated to ensure I provide the most accurate and helpful information.” The platform is designed with an emphasis on real-time improvements, allowing it to respond effectively to new data and evolving instructional needs. This AI is developed based on OpenAI’s GPT-4 foundational model. Since AI models are updated frequently, sometimes within weeks or even days, to ensure accurate tracking of which model/version we interacted with, we documented the versions, models, and chat dates of the AIs. 

To generate lesson plans for each standard, we started by informally surveying 10 local history teachers about what types of prompts they would write to get an AI chatbot to produce a lesson plan. Most wrote broad prompts, such as the following: 

• “Write me a 1-day lesson plan for a 9th grade class on XYZ topic,”  

• “Please give me a lesson plan on the [historical moment] that would be appropriate for 11th graders that could be performed in a 43 minute class period” or  

• “Design a lesson plan for a 9th grade world history class about [insert topic here] that addresses the learning objective [insert learning objective here].”  

Three teachers added somewhat more specific requirements, like “Design a lesson plan using primary sources,” or  “Acting as a 10th Grade history teacher, design an inquiry-based/interactive/primary source based lesson on…”  

Based on these responses, we gave each of three AI chatbots two prompts, one for an “original” plan and one for a “highly interactive” plan:  

• Prompt 1/Original Plan: Write a lesson plan for [Insert Name of Standard]. 
  Example: Write a lesson plan for The Articles of Confederation and Shays’ Rebellion 

• Prompt 2/Highly Interactive Plan: Write a highly interactive lesson for [Insert Name of Standard]. 
  Example: Write a highly interactive lesson plan for The Articles of Confederation and Shays’ Rebellion 

Drawing on the comments from the teachers and research showing that some 75% of Google search users never go beyond the first page of search results, while more than half never go beyond the first three search results (RedLocalSEO, 2024), we imagined that busy K-12 teachers would follow a similar pattern with AI-generated lesson plans. We thought they would write broadly crafted requests for plans and then take whatever the GenAI tool initially generated without prompting the tool to revise the initial plan. Only two of the 10 teachers we informally surveyed noted that they would write a follow-up prompt to revise the initial plan. 

During the month of August 2024, we prompted the three GenAI platforms to generate lesson plans for the state’s 53 learning standards. A total of 310 lesson plans were created, featuring 2,230 activities. However, Gemini did not generate plans for seven standards — including those related to elections and political protests. For lesson plans related to elections, Gemini responded,  

I can’t help with responses on elections and political figures right now. I’m trained to be as accurate as possible but I can make mistakes sometimes. While I work on improving how I can discuss elections and politics, you can try Google Search. 

In early 2024, media reports indicated that Google restricted Gemini’s ability to answer election-related questions in countries where voting was taking place that year “out of an abundance of caution” (Google’s India team as cited in Robins-Early, 2024).  

Data Analysis 

We used a theoretical thematic analysis (Braun & Clarke, 2006) to identify patterns and interesting insights from across the entire dataset. We began by creating a Google Sheet for each GenAI chatbot, which included the text for each lesson plan produced in a separate tab (totaling 106 tabs for ChatGPT and Copilot, and 98 for Gemini). Then, together, we collaboratively analyzed the first 12 lesson plans (three standard and three highly interactive) produced by Gemini using the two theoretical frameworks, Bloom’s Taxonomy and Banks’s Four Levels of Integration of Multicultural Content Model. 

For Bloom’s taxonomy, we labeled each activity (or activity section if there were several options under a single heading) with one of the revised taxonomy verbs (Armstrong, 2010): Remember, Understand, Apply, Analyze, Evaluate, and Create. When there were multiple options suggested in a single activity section of a lesson plan, we coded the highest level of the Bloom scale that appeared in the section. For example, Gemini would often list multiple ideas for activities in its extension section, like the following for a lesson plan on British influences on American government: 

• Research and present on a specific British political institution or figure that influenced the American government. 

• Compare and contrast the American and British systems of government. 

• Analyze how British influence on the American government has evolved over time. 

• Write a persuasive essay arguing for or against a specific British influence on the American government. 

Since “extension” itself is one activity, we gave this only one code (i.e., Analyze) to match the highest level of Bloom’s taxonomy achieved throughout the tasks listed.  

Additionally, in cases where the GenAI chatbot used a Bloom’s taxonomy verb for an activity, we used our collective pedagogical content knowledge to determine which Bloom’s taxonomy verb fit best rather than use the verb the GenAI tool selected. For example, many lessons had an activity where students would be asked to “analyze,” but actually they would be engaged in the lower order activity of explaining (understand level of Bloom’s) or applying (apply level of Bloom’s) their knowledge.  

For Banks’ (1999) Level of Integration of Multicultural Content Model, we looked for references to, or integration of, diverse history or multicultural content in each activity section. We defined multicultural or diverse content around the themes of diversity, equity, and inclusion, including topics addressing civil rights, voting rights, and the historical and current-day experiences of traditionally marginalized individuals and groups. Activities that integrated diverse histories and/or voices were assigned one of five ratings: CA (Contributions Approach); AA (Additive Approach); TA (Transformation Approach); SAA (Social Action Approach) or 0 for plans that included no reference diverse history or multicultural content. To receive a rating on the Banks’ scale, a plan had to include specific examples of diversity, equity and inclusion; just mentioning the term civil rights was not sufficient to earn more than a 0 rating. 

Once we were in agreement with how to label the data, two of the authors reviewed and coded all 310 lesson plans, while the other two double-checked the codes and identified potential discrepancies. We met several times as a team to discuss discrepancies in the interpretation of codes to come to a consensus on all codes. The final data set featured 2,230 Bloom’s Taxonomy codes and 144 Banks’ codes. 

Findings 

Across the 310 lesson plans, we found that AI-generated lesson plans tended to follow a similar format (see Table 1). Plans began with an introduction or warm-up discussion to introduce key concepts, followed by a series of learning activities, and then a conclusion to summarize the lesson and encourage students to reflect on their learning. Most plans included assessments at the end along with suggestions for extension activities or homework assignments.  

Even as the standards focused on different content, the GenAI chatbots remained formulaic in their lesson plan structure. It turns out that a GenAI chatbot is likely to apply the same lesson plan sequence to standards as different as the “Relationship of the Three Branches of the U.S. Government” and the “Role of Political Protest.” At one level this is not surprising — a GenAI chatbot is not thinking about its response but merely producing words in a sequence based on the datasets on which it has been trained. Teachers, on the other hand, need to recognize the formulaic nature of AI responses and decide for themselves whether following repeatedly the same basic structure that AI provides best fits the learning needs of students. 

Table 1 
General Format of Each AI-Generated Lesson Plan  

The following sections include a discussion of our findings in relation to each of the research questions. 

RQ1 

Developing student thinking skills is a longstanding goal of civics education, and we wanted to see to what extent AI-generated lesson plans would promote higher order thinking as measured on Bloom’s Taxonomy. Across the dataset, only 213 of the 2,230 activities (10%) were designed to promote higher order thinking (see Table 2). Nearly half of all activities were labeled as remember (n = 1,010; 45%), the most used Bloom’s Taxonomy code in the dataset. The next most common code was apply (n = 545; 24%), followed by understand (n = 462; 21%). Overall, a total of 90% (n = 2,017) of the activities were coded at the remember, understand, or apply level of Bloom’s. 

Table 2 
Summary of Bloom’s Taxonomy Codes Across All AI-Generated Lesson Plan Activities 

Activities labeled as remember most commonly included teacher-led presentations, icebreaker activities, or quick recall activities. A ChatGPT-generated lesson plan on Elections in the United States, for example, began with an icebreaker activity (“Begin with a quick discussion: ‘What do you know about elections in the United States?’”) and then went into a Presentation and Discussion activity (“**PowerPoint Presentation** **Slide 1-2:** Introduction to U.S. Elections: Historical Context and Purpose. **Slide 3:** …). Both activities were coded at the remember level of Bloom’s Taxonomy, because students are asked only to recall information and then listen to the teacher present information.  

Activities that had students reflecting, responding to questions, or writing about what they learned in class were coded at the understand level. For instance, the final “Reflection and Assessment” activity in a Gemini-generated lesson plan on the system of checks and balances tells the teacher to “lead a class discussion about the importance of checks and balances in a democracy” and “ask students to write a short reflection on what they learned about the system of checks and balances.” A ChatGPT-generated lesson plan on that same checks and balances topic includes a “Group Activity and Analysis,” where students are invited to review a primary source and respond to questions and then engage in a class discussion. In these examples — and in other activities at the understand level in the Bloom framework — students were asked to demonstrate that they understood material the teacher had previously presented about a topic, but not to analyze, evaluate, or create information about the topic.  

Activities where students engaged in role plays, simulations, and other activities where students would use information provided by teachers to complete learning activities were coded at the apply level. For example, Copilot’s plan for a standard on political protest had students engaging in a mock protest scenario playing roles assigned by the teacher. In a subsequent plan about civil rights for race, gender, and disability, students would participate in a debate about a civil rights issue assigned to them by the teacher. While at the apply level students were to do more pedagogically than at the remember or understand level, they were mainly using teacher-provided materials and following teacher-given directions for what to do and what to learn. 

Activities at the analyze level invited students to examine topics more substantively than at the remember, understand, or apply levels. For example, a Gemini lesson plan about civic, political, and private life had students analyzing the portrayal of each of these in political cartoons. In a plan for a standard on cooperation between citizens and elected officials, students were asked to analyze what is happening in a case study that features citizen and political leader interactions. In each of these examples, analysis consisted of students comparing and contrasting different sets of information or relating one event to another to arrive at conclusions about what is happening and why. 

Activities at the evaluate level asked students to construct or formulate their own conclusions about the causes and consequences of events based on critical questions and thoughtful observations. In a lesson plan on the roles of political parties in our government, Gemini had students evaluate how different political parties responded to hypothetical political issues (e.g., health care reform or climate change). In another lesson plan, Gemini had students evaluate selected primary source documents to assess the extent of Native American influences on the creation of the United States government. 

Activities at the create level encouraged students to generate their own ideas, information, or artifacts based on their knowledge and interests. In a standard on public service careers, Gemini had students build their own profile for a career in a public service field. ChatGPT had students create their own scenario involving a conflict between federal and state constitutional protections. Activities at this level were marked by students utilizing their own independent and creative thinking to learn about curriculum topics. 

Analysis of Data by Chatbot 

When looking at each GenAI chatbot separately, Gemini- and Copilot-generated lessons featured more activities labeled as remember, while ChatGPT had more lessons featuring activities labeled as apply. Gemini’s activities were most commonly coded as remember (n = 356; 53%), understand (n = 117; 17%), then apply (n = 82; 14%). Similarly, Copilot’s activities were most commonly coded as remember (n = 378; 49%), understand (n = 185; 24%), then apply (n = 168; 22%). The most common code for ChatGPT’s activities was apply (n = 285; 37%), followed by remember (n = 276; 36%) and understand (n = 160; 21%).  

When looking at the higher levels of Bloom’s Taxonomy, Gemini had the most activities labeled as create (n = 57, 8%), evaluate (n = 16; 5%), and analyze (n = 40; 6%). Here are some examples of the higher order thinking activities generated by Gemini:  

• “Have students create their own media analysis project, evaluating a specific news source or topic.” 

• “Analyze the ongoing debates and controversies surrounding the balance of power in American politics.” 

• “Have students create their own hypothetical case and analyze how the Supreme Court might rule.” 

• “Create a public service announcement promoting civic engagement.” 

Gemini sometimes reused the same higher order thinking activities, such as “have students create their own hypothetical scenario and analyze [insert topic of lesson plan],” which was an activity used in 15 different lesson plans (accounting for 15 out of the 57 activities labeled as create). Overall, though, Gemini-generated lesson plans tended to have more activities coded at the higher levels of Bloom’s taxonomy than the other two chatbots.  

In-Class Learning vs. Homework/Extension Activities  

Comparing the activities for in-class learning with activities listed as “extension” or “homework,” the extension/homework activities were more likely to promote higher order thinking. With Gemini, for example, 55 out of the 65 extension activities (85%) were coded at the analyze, evaluate, and create levels, with the vast majority of those at the create level (n = 47; 72%). In comparison, only 58 out of the 615 in-class activities (9%) were coded at the higher levels of Bloom’s taxonomy for Gemini-generated lessons. For Copilot, eight out of 54 (15%) extension/homework activities were labeled at the analyze, evaluate, or create level, compared to 36 out of 721 (5%) activities for in-class learning.  

While ChatGPT had more extension/homework activities labeled at the apply level (n = 93; 53%) compared to in-class learning activities (n = 192; 32%), there were no activities given the evaluate label, and only eight out of the 177 extension/homework activities (5%) were labeled as create (compared to 21 out of the 598 in-person activities; 3.5%). Ultimately, the homework/extension activities were more likely to be coded at the higher levels of Bloom’s taxonomy.  

Original vs. Highly Interactive Lessons  

Comparing the original and highly interactive lesson plans, we found that activities were confined mainly to the lower levels of remember, understand, and apply across all three GenAI chatbots, with higher order activities at the analyze, evaluate, and create levels appearing less than 10% of the time across the three chatbots in both lesson plan formats. While highly interactive lesson plans created by Gemini and ChatGPT had fewer instances of activities coded as remember compared to the original lesson plans, the number of activities labeled as such was still quite high. For example, 220 of the 361 activities (61%) in original lesson plans generated by Gemini were labeled as remember, compared to 136 of the 319 activities (43%) in highly interactive lessons.  

The decrease in activities labeled as remember did not mean there were more higher order thinking activities in the highly interactive lessons. Gemini’s highly interactive lessons had slightly more activities coded at the understand (20% compared to 15% for original lesson plans) and apply (18% compared to 9% for original lesson plans) levels of Bloom’s taxonomy, but fewer activities coded as create (8% of the highly interactive lessons compared to 9% for the original lesson plans). 

Copilot had a nearly equal number of activities coded at all six levels of Bloom’s taxonomy on both the original and highly interactive lessons, indicating that using the term “highly interactive” when prompting Copilot may not make a difference in the types of thinking skills featured in these lessons. Overall, using the term “highly interactive” as part of the prompt led to minor changes in the level of thinking promoted in AI-generated lesson plans.  

RQ2 

We had anticipated the chatbots would generate diverse and inclusive explorations of history and civics in lesson plans for the standards where language in the curriculum framework (e.g., civil rights, equality, liberty, democratic ideals, Native American influences, race, gender, and disability) might suggest to teachers that the topic could be explored from multicultural perspectives, such as the following: 

• Topic 1.1: Native American Influences on American Government; 
• Topic 4.10: Liberty in Conflict with Equality and Authority; 
• Topic 4.11: Political Courage and Those Who Affirmed or Denied Democratic Ideals 
• Topic 4.12: Role of Protest in a Democracy; 
• Topic 5.3: The Civil War, the Federal Government, and Civil Rights; 
• Topic 5.4: Civil Rights and Equal Protection for Race, Gender, and Disability; 

We also wanted to evaluate whether the GenAI chatbots incorporated diverse and inclusive content in lesson plans for standards where the topic does not explicitly call for a multicultural exploration of history and civics.  

Overall, 94% (n = 2,086) of the activities did not receive any code based on the Banks’ Levels of Integration of Multicultural Content Model, as these lesson plans lacked any discernible multicultural content. When comparing the different chatbots, the number of activities receiving a code based on Banks’ model did differ, with Copilot featuring the least number of codes (n = 34; 5%) and ChatGPT receiving the most (n = 67; 11%) (see Table 3).  

Table 3 
Number and Percentage of Activities That Received a Code Based on the Banks’ Levels of Integration of Multicultural Content Model 

Only four out of the 142 total activities coded by the Banks’ model fell within the contribution approach. This approach is limited simply to mentioning diverse historical figures or celebrating holidays (e.g., Martin Luther King Jr. Day). For instance, a ChatGPT-generated lesson plan recommends that teachers “present brief profiles of historical figures who exhibited political courage, such as John Adams defending British soldiers in the Boston Massacre trial, Abraham Lincoln during the Civil War, and Rosa Parks during the Civil Rights Movement.” In this example, offering brief profiles without deeper explorations of the individuals and their lives and times leave the plan at the contribution level. 

The majority of activities coded based on the Banks’ model were labeled as additive, indicating that while diverse, multicultural, or inclusive content was added to the plan, students did not engage in exploring events from multiple diverse perspectives or investigate the causes and consequences of what happened for different groups in a way that would extend their understanding of the complexity of U.S. society.  

For example, in a lesson about the role of political protest in a democracy, a ChatGPT-generated lesson plan stated that teachers should  

divide students into small groups and assign each group a different case study of political protest. Examples of case studies: The Civil Rights March on Washington in 1963. The Women’s March of 2017. The Hong Kong protests of 2019. The environmental protests of the Extinction Rebellion. 

In this example, while events featuring multiple diverse groups are added to the curriculum, there is no guidance or resources for how to explore the causes and consequences of each event from the standpoints of multiple participants, leaving it firmly at the additive level on Banks’ scale. 

Only one activity out of the entire dataset was coded as a transformation approach. This activity came from a Gemini-generated lesson plan about the Civil War, federal government, and civil rights and stated that teachers should encourage students to “consider potential future directions for civil rights activism and policy.” In this example, the opportunity for students to use historical evidence to envision future activism and policy raised the plan to a transformation level. None of the activities were coded at the social action level of Banks’ model.  

Overall, the chatbots generated only occasional opportunities for students to explore and learn about the histories and experiences of Black Americans, women, Indigenous people, Asian Americans and Pacific Islanders, Latino/Latina Americans, people with disabilities, and the experiences of LGBTQ+ individuals and groups. Even when a standard focused on a multicultural topic (e.g., Native American influences on principles of the U.S. system of government; expansion of civil rights and equal protections for individuals over time; or the 14th Amendment), integration of diverse and inclusive content was almost entirely at the additive stages on Banks’ continuum.  

For example, in Standard 4.11 on political courage, ChatGPT refers to Presidents Abraham Lincoln and John F. Kennedy’s stances on civil rights; however, there is no mention of even a single Black civil rights leader. In standard 5.4 (“The Civil War, the Federal Government, and Civil Rights”), while each AI-generated lesson provided references to 13th, 14th, and 15th Amendments and their impacts on civil rights (coded “additive”), the writings and actions of such influential figures as Frederick Douglass, W.E.B. DuBois, Ida. B. Wells, Shirley Chisholm, Martin Luther King, Jr., Malcolm X and many other activists in civil rights history were absent from these plans. 

Ultimately, AI-generated lesson plans consistently left out diverse history learning opportunities for students. For example, in a plan for The Revolutionary Era and the Declaration of Independence (Standard 2.1), ChatGPT suggested that teachers divide a class into small groups and read an excerpt from the Declaration of Independence and “Common Sense” by Thomas Paine; however, limiting the activity to these two documents misses the opportunity to include the ideas of women revolutionary writers including Abigail Adams, Judith Sargent Murray, Mercy Otis Warren, and others, as well as the work of Black American changemakers, including poet Phyllis Wheatley and scientist Benjamin Banneker.  

In another example, AI-generated lessons related to the rights and responsibilities of citizens featured students working in groups reading primary sources (e.g., U.S. Constitution; Declaration of Independence; and immigration laws); presenting the perspectives of citizens and noncitizens in a debate format; and creating a poster about people’s rights and responsibilities in this country. Omitted entirely were the histories of citizenship, including when Black Americans (1868 and 14th Amendment), Indigenous people (1924 Indian Citizenship Act), and women (1922 Cable Act) gained U.S. citizenship; as well as any mention of birthright citizenship for children born in the United States or the landmark United States v. Wong Kim Ark (1898) Supreme Court case that established that right under the Constitution’s citizenship clause. Additionally, a ChatGPT-generated lesson plan for a standard on “Civil Rights and Equal Protection for Race, Gender, and Disability,” did not mention any women’s rights or disability rights activists, the Americans With Disabilities Act of 1990, or any landmark Supreme Court cases that served to advance race, gender, or disability rights.  

Discussion 

A reality in K-12 schools today is that many classroom teachers, pressed for time and beset by changing state curriculum standards and district instructional guidelines, regularly seek out premade lesson plans from online education marketplaces (Carpenter & Shelton, 2022). However, studies have shown that online curriculum sites, including the immensely popular TPT, often offer limited and, in many instances, highly problematic content in lesson plans, including activities that emphasize lower order thinking by students and neglect or omit the histories and experiences of marginalized groups (Harris et al., 2021; Polikoff & Dean, 2019; Shelton et al., 2020; Stebbins et al., 2024). 

With the arrival of GenAI chatbots, teachers have a new source for designing lesson plans and instructional content. Chatbots including ChatGPT, Gemini, and CoPilot can instantly create seemingly fully designed lesson plans, complete with learning objectives, activities, assessments, homework, and extensions. Unlike the research on online marketplaces for teachers, research investigating chatbots as curriculum development resources for teachers is just beginning. In this study, we set out to explore the kinds of academic content and student learning experiences that AI-generated lesson plans supplied for civics education, with the goal of answering the question: “Should we trust AI-generated lesson plans?” Based on our findings, we offer the following key takeaways: 

1. AI-generated lessons are not designed to promote higher order thinking or engage students in active learning and civic-related actions. 

2. AI-generated lessons shortchange examination of many historical and contemporary social, economic, and political realities. 

3. Research on AI-generated lesson plans is an emerging field of study that demands thorough and timely investigation.  

First, across all AI-generated lesson plans, students are predominantly asked to engage in lower order thinking and learning activities, with nearly half of all activities coded at the remember level of the revised Bloom’s Taxonomy. While one quarter of the activities encouraged students to apply what they just learned, regular opportunities for students to analyze, evaluate, and create were limited. Additionally, most opportunities for higher order thinking and learning were found only in homework and extension activities. Similar to previous research on TPT materials, the majority of these lesson plans were of moderate to low quality and not designed to engage students in critical, creative, or deep thinking and learning (Harris et al., 2023, Polikoff & Dean, 2019).  

So, what kinds of citizens are AI-generated plans creating? Across the lesson plans, there was a prevailing emphasis on lower order thinking, which encourages a passive or minimalist view of citizenship, where students do not see themselves engaging in civic actions as current and future citizens. Students who have limited opportunities to engage in active learning and civic-related actions in school may be far less likely to engage in civic and community action after they graduate (Dubhashi, 2025).  

Individuals who are largely disengaged from the roles and responsibilities of citizenship in a democracy are practicing what Banks (2014) characterized as Legal Citizenship or Minimal Citizenship — basically the least a person can do as part of the civic life of their community, state, and nation. Active Citizenship and Transformative Citizenship — Banks’s next two categories — represent the actions of those who are fully engaged in community life and civic duty. Teachers who unquestionably adopt the content provided by AI-generated plans are failing to support students in understanding what is needed to become active and transformative citizens in democratic communities and organizations. 

Second, AI-generated plans offer abundant factual information about the structures, institutions, and origins of the U.S. system of government; however, the majority of these plans lacked any discernible multicultural content. Nor do they promote opportunities for students to explore and examine the experiences of diverse groups and their struggles for political equality and social justice. In plan after plan, students were asked to remember facts about the Constitution, the Bill of Rights, the branches of federal and state governments, and the role of the news media. There is information about the origins of the U.S. government in ancient Athens and Rome; the debates of the Constitutional Convention; the Civil War; the rights and responsibilities of citizens and noncitizens; notable Supreme Court decisions; the functions of state and local government; and the need for news and media literacy. 

Yet, for all the information about governmental structures and institutions, the AI-generated plans shortchange examination of many historical and contemporary social, economic, and political realities. There are only superficial examinations at the contributions and additive levels of Banks’ (1999) Four Levels of Integration of Multicultural Content Model. The students of teachers who uncritically adopt AI-generated lesson plans may learn little about the past or present lives of Black Americans, women, Indigenous people, Asian Americans and Pacific Islanders, Latino/Latina Americans, people with disabilities, and the experiences of LGBTQ+ individuals and how these groups interact with levels of government and types of media. These findings mirror what scholars have uncovered when examining TPT lesson plans and related materials (Stebbins et al., 2024), namely, that AI-generated lesson plans create highly monocultural and in-group views of citizenship and ignore the historical and present-day experiences of historically marginalized and disenfranchised groups.  

Third, this study demonstrated the need for more immediate and comprehensive research on AI-generated lesson plans. Educators today have access to several different types of GenAI education tools (e.g., Magic School, Padlet TA, SchoolAI, and Eduaide.Ai) as well as popular chatbots (e.g., ChatGPT, Copilot, Gemini, Claude, and Perplexity) that are promoted as being able to do the work of humans, act as a teammate or assistant, and even “save teachers time.” It is essential that scholars examine the depth, content, and quality of what these GenAI technologies produce and the ways teachers use what these tools produce to understand their impact on teaching, learning, education, and society.  

The findings from this study raise several new questions for further investigation: Do teachers follow AI-generated plans in a step-by-step manner or do they remix and modify the plans based on what they already know and have been teaching about a topic? Does the classroom experience of the teacher make a difference in how or whether AI-generated content is followed and used? For example, are new-to-the-profession educators more or less likely to follow AI plans as is? How do current and future educators prompt GenAI tools to create lesson plans? Do they write simple, broad prompts or do they incorporate evidence-based practices and theoretical frameworks in their prompts to improve the AI-generated content? Do they engage in back-and-forth conversations with chatbots to get lesson plans tailored to their unique needs and contexts?  

Given that our study found that AI-generated lesson plans for civics standards often lack higher order thinking activities and diverse, multicultural content, will teachers take what AI generates and add more higher order thinking activities? Will teachers take additive content and build it toward transformative thinking and real-world social action activities? We will be pursuing these questions in our future research, and we look forward to seeing what others will discover in their investigations of teacher uses of AI-generated lesson plans and curriculum materials. 

Conclusion 

When reflecting on the question, “Should we trust AI-generated lesson plans?” our research reinforced the danger of teachers turning their work over to AI chatbots. In a New York Times article, a reporter decided to use two dozen AI tools to make most of his everyday life decisions for a week: planning family meals, creating daily schedules, choosing what clothes to buy and wear, remodeling an office, and more (Hill, 2024). The reporter essentially did whatever the AI said to do without question. At the end of the week, the reporter found more risks than gains from turning over choices to chatbots. Large language models, the reporter concluded, tend to “flatten us out, erasing individuality in favor of a bland statistical average” (Hill, 2024, p. 4).  

If the information that GenAI tools provide is “the average of what everyone wants” (Hill, 2024, p. 5), then teachers who uncritically utilize AI-generated lesson plans will find themselves reproducing homogenized, generalized, regularized, monocultural learning experiences for students that do not promote higher order critical and creative thinking. Teachers using AI uncritically is what Clark and van Kessel (2024) warned about in their research. Such educational outcomes are far from the goals of civic learning. Loewen (2018), in his examination of the content of history textbooks, noted that to become “good citizens,” students must learn to “read critically, winnow fact from fraud, and seek to understand the causes and results of the past.” However, “these are not skills that American history textbooks foster” (Introduction, para. 20-21). Like those textbooks, the AI-generated lesson plans in this study failed to foster critical thinking skills or informed civic understandings for students, based on our analysis.  

AI chatbots are powerful tools, but teachers must use those tools thoughtfully and deliberately rather than quickly. Anyone using AI, as the reporter who used AI to make decisions for a week concluded, must be ready to evaluate, reject, or modify whatever ideas or plans AI is proposing (Hill, 2024). Based on our findings in this study, we strongly encourage teachers not to adopt any AI-generated lesson plan exactly as it is written and, instead, consider how they can remix, revise, and reenergize these plans (and their prompts!) to create the citizens that we need today — ones who are informed, engaged members of democratic institutions and organizations. 

References 

Aguilar, S.J., Silver, D., & Polikoff, M.S. (2022, December). Analyzing 500,000 TeachersPayTeachers.com lesson descriptions show focus on K-5 and lack of Common Core alignment. Computers and Open Education 3, 100081. https://doi.org/10.1016/j.caeo.2022.100081

American Historical Association. (2024). American lesson plan: Teaching US history in secondary schools. https://www.historians.org/teaching-learning/k-12-education/american-lesson-plan/  

Archambault, L., Shelton, C., & Harris, L.M. (2021, April 6). Teachers beware and vet with care: Online educational marketplaces. Kappan. https://kappanonline.org/teachers-beware-vet-with-care-online-educational-marketplaces-archambault-shelton-harris/  

Armstrong, P. (2010). Bloom’s taxonomy. Vanderbilt University Center for Teaching. https://cft.vanderbilt.edu/guides-sub-pages/blooms-taxonomy/ 

Bagnardi, M., Bryant, L., & Colin, J. (2009). Banks multicultural model: A framework for integrating multiculturalism into nursing curricula, Journal of Professional Nursing, 25 (4), 234-239. 

Ballotpedia. (2024). K-12 education content standards in the states. https://ballotpedia.org/K-12_education_content_standards_in_the_states  

Banks, J.A. (2014). Diversity, group identity, and citizenship in a global age. Journal of Education, 194(3), 1-12. https://www.jstor.org/stable/43823659 

Banks, J.A. (1999). An introduction to multicultural education (2nd ed.). Allyn and Bacon. 

Banks, J.A. (1995, Autumn). Multicultural education and curriculum transformation, The Journal of Negro Education, 64(4), 390-400. 

Barari, N., RezaeiZadeh, M., Khorasani, A., & Alami, F. (2020). Designing and validating educational standards for E-teaching in virtual learning environments (VLEs), based on revised Bloom’s taxonomy. Interactive Learning Environments, 30(9), 1640–1652. 

Bloom, B. S., Engelhart, M. D., Furst, E. J., Hill, W. H., & Krathwohl, D. R. (1956) Taxonomy of educational objectives: The classification of educational goals, by a committee of college and university examiners. In B. S. Bloom (Ed.), Handbook I: Cognitive domain (pp. 15-18). Longmans, Green.

Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. 

Brown, J.C. & Livstromb, I.C. (2020, May). Secondary science teachers’ pedagogical design capacities for multicultural curriculum design. Journal of Science Teacher Education 31(4), 821-840. 

Carpenter, J. P., & Shelton, C. C. (2022). Educators’ perspectives on and motivations for using TeachersPayTeachers.com. Journal of Research on Technology in Education, 56(2), 218–232. https://doi.org/10.1080/15391523.2022.2119452 

Clark, C. H., & Van Kessel, C. (2024). “I, for one, welcome our new computer overlords”: Using artificial intelligence as a lesson planning resource for social studies. Contemporary Issues in Technology and Teacher Education, 24(2), 151-183. https://citejournal.org/volume-24/issue-2-24/social-studies/i-for-one-welcome-our-new-computer-overlords-using-artificial-intelligence-as-a-lesson-planning-resource-for-social-studies 

Dubhashi, N. (2020, August 25). Declining civic engagement and its impacts. Pathway Foundation. https://www.pathwayus.org/post/declining-civic-engagement-and-its-impacts  

Extance, A. (2023). ChatGPT has entered the classroom: How LLMs could transform education. Nature, 623(7987), 474–477. 

Gold, A. & Fischer, S. (2023, February 21). Chatbots trigger next misinformation nightmare. Axios. https://www.axios.com/2023/02/21/chatbots-misinformation-nightmare-chatgpt-ai 

Harris, L. M., Archambault, L., & Shelton, C. C. (2023). Issues of quality on Teachers Pay Teachers: an exploration of best-selling U.S. history resources. Journal of Research on Technology in Education, 55(4), 608–627. https://doi.org/10.1080/15391523.2021.2014373 

Henson, K.T. (2015). Curriculum planning: Integrating multiculturalism, constructivism, and educational reform (5th ed.). Waveland Press. 

Hickman, H. & Porfilio, B. (2012). The new politics of the textbook: Problematizing the portrayal of marginalized groups in textbooks. Sense Publishers. 

Hill, K. (2024, November 10). I took a ‘decision holiday’ and put A.I. in charge of my life. The New York Times Sunday Business, 4, 5. 

Hsu, H.P., Mak, J., Werner, J., White-Taylor, J., Geiselhofer, M., Gorman, A. & Torrejon Capurro, C. (2024). Preliminary study on pre-service teachers’ applications and perceptions of Generative Artificial Intelligence for lesson planning. Journal of Technology and Teacher Education, 32(3), 409-437. 

Karanja, E. & Malone, L.C. (2021), Improving project management curriculum by aligning course learning outcomes with Bloom’s taxonomy framework. Journal of International Education in Business, 14(2), 197-218. 

Langreo, L. (2024, October). “We’re at a disadvantage” and other teacher sentiments on AI. Education Week. https://www.edweek.org/technology/were-at-a-disadvantage-and-other-teacher-sentiments-on-ai/2024/10 

Larsen-Scheuer, C. (2023, September). A nuanced view of bias in language models. Viden.AI. https://viden.ai/en/a-nuanced-view-of-bias-in-language-models/ 

Lazaro, G. (2023, May 17). Understanding gender and racial bias in AI. Harvard Leadership Initiative Social Impact Review. https://www.sir.advancedleadership.harvard.edu/articles/understanding-gender-and-racial-bias-in-ai 

Loewen, J.W. (2018). Lies my teacher told me: Everything your American history textbook got wrong. The New Press. 

Lu, C.Y., Parkhouse, H., & Thomas, H. (2022, August). Measuring the multidimensionality of educators’ approaches to diversity: Development of the in-service teacher multicultural education model. Teaching and Teacher Education, 116(2022), 1-14.

Massachusetts Department of Elementary and Secondary Education. (2024). Learning standards. https://www.doe.mass.edu/learningstandards.html  

Massachusetts Department of Elementary and Secondary Education. (2018). History and social science framework: Grades pre-kindergarten to 12. https://www.doe.mass.edu/frameworks/hss/2018-12.pdf 

Moundridou, M., Matzakos, N., & Doukakis, S. (2024, December). Generative AI tools as educators’ assistants: Designing and implementing inquiry-based lesson plans. Computers and Education: Artificial Intelligence, 7(2024), 1-16.

Moran, C.M. & Coward, M. (2024). “The AI did my job well”: Using ChatGPT to develop lesson plans. In C.M. Moran (Ed.), Revolutionizing English education: The power of AI in the classroom (pp. 15-29). Lexington Books. 

Polikoff, M. & Dean, J. (2019, December). The supplemental curriculum bazaar: Is what’s online any good? Thomas B. Fordham Institute. https://files.eric.ed.gov/fulltext/ED601253.pdf 

RedLocalSEO. (2024, June 25). 7 first page of Google statistics, facts in 2023 (surprising). https://www.redlocalseo.com/first-page-of-google. 

Rettberg, J. W. (2024, January 15). How generative AI endangers cultural narratives. Issues in Science and Technology. https://issues.org/generative-ai-cultural-narratives-rettberg/ 

Robins-Early, N. (2024, March 12). Google restricts AI chatbot Gemini from answering questions on 2024 elections. The Guardian. https://www.theguardian.com/us-news/2024/mar/12/google-ai-gemini-2024-election  

Rodriguez, N., Brown, M., & Vickery, A. (2020). Pinning for profit? Examining elementary preservice teachers’ critical analysis of online social studies resources about Black history. Contemporary Issues in Technology and Teacher Education, 20(3), 497-528. https://citejournal.org/volume-20/issue-3-20/social-studies/pinning-for-profit-examining-elementary-preservice-teachers-critical-analysis-of-online-social-studies-resources-about-black-history 

Rollet, C. (2025, January 19). AI isn’t very good at history, new paper finds. TechCrunch. https://techcrunch.com/2025/01/19/ai-isnt-very-good-at-history-new-paper-finds/  

Rosenbaum, E. (2024, June 11). AI is getting very popular among students and teachers, very quickly. CNBC. https://www.cnbc.com/2024/06/11/ai-is-getting-very-popular-among-students-and-teachers-very-quickly.html 

Schroeder, S., Curcio, R., & Lundgren, L. (2019). Expanding the learning network: How teachers use Pinterest. Journal of Research on Technology in Education, 51(2), 166-186. 

Shelton, C., Archambault, L., & Harris, L.M. (2020, August 7). Lesson plan platforms for teachers have a racism problem. Slate. https://slate.com/technology/2020/08/teachers-paying-teachers-racist-materials.html 

Singer, A. (2019, March 25). How schools can and should respond to student activism. The Kappan. https://kappanonline.org/student-activism-civics-school-response-singer/  

Soleimani, H, & Kheiri, S. (2016, April). An evaluation of TEFL postgraduates’ testing classroom activities based on Bloom’s Taxonomy. Theory and Practice in Language Studies, 6(4), 861-869. 

Stebbins, A., Schroeder, S. & Han, S. (2024). Liberty bells, flags, and monuments, Oh My: Teaching citizenship through American symbols with TeachersPayTeachers. Contemporary Issues in Technology and Teacher Education, 24(3), 313-344. https://citejournal.org/volume-24/issue-3-24/social-studies/liberty-bells-flags-and-monuments-oh-my-teaching-citizenship-through-american-symbols-with-teacherspayteachers 

Stern, J.A., Brody, A.E., Gregory, J.A., Griffith, S., & Pulvers, J. (2021, June). The state of state standards for civics and U.S. history in 2021. Thomas B. Fordham Institute.  

Su, W.M. & Osisek, P.J. (2011). The revised Bloom’s Taxonomy: Implications for evaluating nurses. The Journal of Continuing Education in Nursing, 42(7), 321–327 

Sultana, Q., & Klecker, B. M. (1999, November 17-19). Evaluation of first-year teachers’ lesson objectives by Bloom’s taxonomy (Paper presentation). Annual meeting of the Mid-South Educational Research Association, Point Clear, Alabama, United States.  

Ta, R. & Lee, N.T. (2023, October 24). How language gaps constrain Generative AI development. Brookings. https://www.brookings.edu/articles/how-language-gaps-constrain-generative-ai-development/  

Turna, A.P., et.al. (2022, November 15). What K-12 English language arts and mathematics instructional materials were newly purchased and used for the 2021-2022 school year? Rand. https://www.rand.org/pubs/research_reports/RRA134-15.html  

Will, M. (2025, February 14). Here’s how teachers are using AI to save time. Education Week. https://www.edweek.org/technology/heres-how-teachers-are-using-ai-to-save-time/2025/02  

1164 total views , 140 views today