From Tool to Thinking: Cultivating Ethical and Analytical AI Use in Global Political Science
James Randle, LaGuardia Community College
Bio
James Randle is a Social Studies teacher at the Academy of American Studies High School with 25 years of experience and a College Now Adjunct Lecturer in the Humanities and Social Sciences Department at LaGuardia Community College. He was a CUNY K–16 Fellow during the 2023–2024 school year. His current focus is on strengthening critical thinking, ethical AI integration, and college readiness in history and political science classrooms while bridging secondary and higher education.
Course Setting
I teach at the Academy of American Studies High School through the College Now partnership with LaGuardia Community College. The course, SSP200 Global Politics, is a college level political science class offered to high school students for dual credit. The class is composed primarily of eleventh graders who represent a range of academic abilities and postsecondary goals. Students engage in rigorous analysis of global political systems, international relations, and comparative government while developing college level reading, writing, and critical thinking skills.
Problem of Practice
How can I use AI to shift students from being 'passive writers' (letting AI do the work) to 'active investigators' (using AI to find facts they’ve already curated)? How can I design a research process where the 'learning' happens during the interaction with the AI, rather than just in the final paper? The current educational climate is fraught with anxiety about AI-assisted cheating. This PoP reframes AI as a “Socratic Partner” rather than an “Answer Machine.” By making the interactions about prompting, probing, and fact-checking, and making that process a visible part of the grade, the project removes the incentive to bypass thinking. It creates a research process where the student’s expertise is demonstrated by how well they direct and verify the AI’s output.Strategy
I implemented a scaffolded AI integration strategy designed to position AI as a research support tool rather than a shortcut.
First, I established clear course norms: AI could be used for brainstorming, outlining, and refining research questions, but students had to document how they used it in a brief AI use statement. I introduced prompt writing through guided practice, showing students how adding context, constraints, and role framing improved output quality. Students revised weak prompts into stronger, more focused ones tied to Global Politics topics.
Next, students created a customized “AI research assistant” by writing a detailed persona prompt aligned with their research question. They used this assistant to generate summaries, identify themes, and develop potential arguments. Draft assignments required students to explain how AI contributed to their thinking and what they chose to accept, modify, or reject.
The rationale draws on research on metacognition and formative scaffolding: when students make their thinking visible and reflect on tool use, they develop stronger self regulation and critical evaluation skills. By integrating AI transparently and reflectively, the strategy reframes it as a cognitive support tool that enhances disciplinary reasoning rather than replacing it.
Documentation
AI Research Assistant Persona Prompt – Human Rights Focus
Sample student work for designing their AI Research Assistant.
AI Use Reflection & Process Log
Sample student entry in AI Journal.
Measuring Impact
Pre- and Post-Student Perception Survey Data
This survey compares student attitudes about AI use before and after the unit. Results show a shift from viewing AI primarily as a shortcut to describing it as a brainstorming and organizational tool. Data demonstrates increased comfort with transparent, ethical AI use in academic work.
Prompt Quality Rubric Scores
This dataset shows rubric scores for student created AI research assistant prompts. Compared to early drafts, final submissions demonstrate increased specificity, clearer constraints, and stronger alignment with global politics concepts. The data reflects measurable growth in prompt writing sophistication and analytical focus.
Analysis
The strategy was largely successful in addressing my problem of practice: encouraging students to engage with AI as a legitimate academic tool rather than a shortcut. By requiring students to design their own AI research assistant and submit a brief reflection on how they used it, AI use became more transparent and purposeful.
Most students moved beyond simple summary prompts and began asking more analytical questions connected to course concepts like sovereignty, power, and human rights norms. I could see evidence of this shift in their drafts, which showed stronger thesis statements and more structured arguments compared to earlier assignments.
Student reflections and informal feedback suggested that the prompt writing workshops were especially impactful. Several students noted that they had previously used AI mainly to “get answers,” but now saw it as a brainstorming and organizing tool. Formative check-ins revealed that some students initially relied too heavily on AI generated language. In response, I added clearer modeling of how to revise AI output and emphasize student voice. This adjustment improved the originality and clarity of later drafts.
Challenges remain. A few students still struggled to critically evaluate AI generated information, and time constraints limited deeper source analysis. In the future, I would incorporate more structured peer review of AI use and dedicate time to explicitly teaching verification strategies. Overall, however, the strategy meaningfully shifted classroom culture toward responsible and reflective AI engagement.
Recommendation
My advice to other professors is to frame AI as a structured academic tool, not a free-for-all resource or something to ban outright. This strategy works best in research based courses where students must analyze complex issues, such as politics, history, or sociology.
Begin with clear norms about acceptable AI use and require transparency through short reflections or process logs. Teach prompt writing explicitly and model how to revise AI generated content so students understand that the thinking still belongs to them.
This is especially effective with high school students in college level courses because it builds metacognitive skills and academic confidence. Many students already use AI informally; guiding that use helps align it with college expectations.
Start small, perhaps with one scaffolded assignment, before expanding. Most importantly, connect AI use directly to disciplinary thinking, so students see it as a tool for deepening analysis rather than replacing effort.