---
source_url: https://citejournal.org/proofing/civic-education-in-the-age-of-ai-should-we-trust-ai-generated-lesson-plans/
ingested: 2026-05-09
sha256: a0b53dc8c431e97ef812543226b4b1b9cda3ba8a1f89f9ac19ef45038cb49e2a
---

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

**Authors:** Trust, T., Maloy, R., Xu, C., & Pelletier, K. (2025)
**Journal:** Contemporary Issues in Technology and Teacher Education, 25(3)
**Source:** https://citejournal.org/proofing/civic-education-in-the-age-of-ai-should-we-trust-ai-generated-lesson-plans/

## Abstract

This study evaluates 310 AI-generated lesson plans (2,230 activities) created by ChatGPT, Gemini, and Copilot for the 53 Massachusetts eighth-grade civics standards. Using Bloom's Taxonomy and Banks' Four Levels of Integration of Multicultural Content, the analysis reveals:

- **90% of activities** promoted only lower-order thinking (remember, understand, apply).
- **94% of activities** lacked any multicultural content; only one activity reached the "transformation" level, and none reached "social action."
- The plans were formulaic, monocultural, and rarely fostered critical or creative thinking.

Recommendation: Teachers must critically evaluate and revise AI-generated lesson plans to provide more expansive, creative, and critical learning experiences.

## Research Questions

- **RQ1:** What type of student thinking skills are promoted in AI-generated civics lesson plans?
- **RQ2:** In what ways do these plans integrate multicultural, diverse, and inclusive perspectives?

## Methodology

- **Data:** 310 lesson plans generated by ChatGPT (GPT-4o), Gemini (1.5 Flash), and Copilot (GPT-4 based) in August 2024.
- **Prompts:** Two per standard:
  - *Original:* "Write a lesson plan for [Standard Name]."
  - *Highly Interactive:* "Write a highly interactive lesson for [Standard Name]."
- **Analysis:** Theoretical thematic analysis using:
  - **Bloom's Revised Taxonomy** (Remember, Understand, Apply, Analyze, Evaluate, Create) — each activity coded at the highest level present.
  - **Banks' Four Levels of Integration of Multicultural Content** (Contributions, Additive, Transformation, Social Action) — coded only if specific diverse content was included.

## Key Findings

### Bloom's Taxonomy Results (RQ1)

| Bloom's Level | Total Activities | % of 2,230 |
|---------------|-----------------|------------|
| Remember      | 1,010           | 45%        |
| Understand    | 462             | 21%        |
| Apply         | 545             | 24%        |
| Analyze       | 85              | 4%         |
| Evaluate      | 31              | 2%         |
| Create        | 98              | 4%         |

- **90% of all activities** at Remember, Understand, or Apply levels.
- "Highly interactive" prompts made minimal difference.
- By chatbot: Gemini had most higher-order (8% Create, 5% Evaluate, 6% Analyze) but still 53% Remember. ChatGPT leaned Apply (37%) + Remember (36%). Copilot heavily Remember (49%) + Understand (24%).

### Banks' Multicultural Content Results (RQ2)

| Banks' Level       | Total |
|--------------------|-------|
| Contributions      | 4     |
| Additive           | 137   |
| Transformation     | 1     |
| Social Action      | 0     |
| No Code            | 2,086 |

- **94% of activities** received no Banks code — no discernible multicultural content.
- Only one activity (Gemini, on civil rights) reached Transformation; none reached Social Action.
- Even standards explicitly mentioning civil rights, Native American influences, or equality were handled at only the Additive level.

### Formulaic Lesson Structure

All chatbots produced nearly identical formats regardless of standard topic: Introduction → Activity 1-4 → Conclusion/Reflection → Assessment → Extension → Homework. The same sequence applied uniformly, resulting in homogenized, non-creative plans.

## Implications

- AI-generated lesson plans, without teacher revision, promote passive learning and moncultural perspectives.
- Teachers must critically evaluate and redesign AI outputs to foster higher-order thinking and multicultural content.
- The study provides concrete evidence for why AI literacy for teachers is essential.