AI Policy Guidance for Education Gem (opens in new tab)

Build with AI
Ready-to-use prompts, Gems, and frameworks for designing transparent, human-centered AI learning experiences.
AI Transparency Booster
Teachers can incorporate this structured transparency prompt into AI-enhanced (AI+) assignments by requiring students to engage with an AI tool (such as Grok, ChatGPT, or similar) for research, brainstorming, drafting, or refinement as part of the task. At the conclusion of their AI session, students simply copy-paste the provided prompt into the same chat thread, which generates a clean, standardized summary including interaction counts, engagement with Costa’s Levels of Questioning (1. Gathering/Input for basic recall; 2. Processing for analysis and inference; 3. Applying/Output for synthesis and creation), an overall usage rating, and positive reinforcement on their strengths.
Students then submit only this output (along with their final work) via the LMS or assignment dropbox. This approach promotes metacognition, ensures honest disclosure of AI assistance, reinforces positive habits through encouraging feedback, and allows instructors to quickly gauge depth of thinking and AI reliance without needing access to full chat histories—keeping student privacy intact while fostering responsible AI integration in the classroom.
If your institution uses Google, you can direct students to use this Google Gem that will ask them to key in “Done” when they are finished to receive the AI Transparency Analysis.
Student Directions: When you finish working with AI on this assignment, paste the prompt below into the same chat. Submit the AI’s full response with your assignment.
Copy and paste prompt for students to use after they engage with AI
Create an honest AI transparency statement for this chat conversation. Analyze only the messages before this one. Base every judgment on evidence from my actual messages. Do not inflate or soften your assessment. An accurate report is more useful to me than a flattering one.
Use these exact sections:
1. Interaction Count: Number of messages I sent.
2. Bloom's Taxonomy Levels: Identify the Bloom's levels my own thinking demonstrated and note the highest. For each level, give one example from my messages and briefly explain why it fits that level. Credit a level only if I did that thinking, not the AI.
3. Who Drove the Thinking: Choose one: Student-Led / Shared / AI-Led. In 2-3 sentences, explain whether I contributed my own ideas, questioned or pushed back on responses, and revised the AI's output, or whether I mainly requested answers and accepted them.
4. Level of AI Use: Choose one: Minimal / Moderate / Substantial / Primary (Primary means the AI produced most of the core work). In 1-2 sentences, name what the AI was used for (brainstorming, explaining, outlining, drafting, editing, fact-checking, etc.).
5. Honest Feedback: Give one specific strength in how I used AI and one specific way I could do more of the thinking myself next time. If my participation was minimal, state that plainly.
Output only these five sections. No introduction, closing, or extra commentary.
See it in action
Below is an example of how the AI Transparency Booster statement worked with my paper “The Flowers, the Flying Machine, and the Genie (opens in new tab).”
Number of messages you sent: 19
You supplied the original ideas, the argumentative direction, the voice specification, and every substantive revision decision; the AI functioned as a drafting partner, research verifier (the 1903 New York Times editorial, the specification-gaming attribution), formatting engine, and editorial sounding board. The AI produced most of the prose in the finished document, but did so against analogies, a thesis, and a set of constraints that originated with you, and you rejected, redirected, or corrected the output at every stage. Rating: Substantial.
Design with AI
This is a curated environment designed to help you navigate the intersection of policy and practice. Whether you are drafting institutional guidance or building classroom assignments, these resources—including the NCCCS AI Guidance Handbook and my TrAIT, AI Sandwich, and AI Scholar frameworks—are here to serve as your foundation.
I invite you to use these starter prompts, or the Google Gemini Gems for educators on this page, to ask questions, brainstorm frameworks, or get started immediately on your own AI design projects.
Note: Users will need to log into their Google account before using the Gemini Gems.
For Policy Design
Act as a policy consultant. Based on the guardrails in the NCCCS Handbook, draft a 'Purpose Statement' and three specific policy guidelines for my department that balance AI innovation with academic integrity.
For Assignment Creation
I want to redesign a writing assignment using the AI Sandwich method and the TrAIT Framework. How can I structure the 'Human-First' and 'Human-Last' steps to ensure students are still meeting the core learning objectives?
For Classroom Transparency (TrAIT)
I am teaching a course that I want to designate as 'AI-Enhanced' under the TrAIT framework. Please analyze the framework's definition of 'AI+' and generate a syllabus policy that encourages students to use tools like ChatGPT for brainstorming and outlining, but strictly prohibits them for final drafting. How should I phrase this to ensure clarity?
For Course Redesign for AI Scholars
I want to upgrade my current course to produce 'AI Scholars.' Based on the framework, give me one example of a classroom activity for each stage of the progression: one that ensures Compliance, one that builds Literacy, and a capstone project that demonstrates Fluency.
Custom assistants
Remember: Think of these Google Gems as a specialized research assistant trained on our specific educational frameworks. They are designed to spark ideas, not dictate final decisions. We strongly encourage you to “push back”—ask follow-up questions, request different tones, or refine the details until it matches your needs.

When AI does the work, students do not learn. This article introduces STEP (Show, Try, Explain, Prove), a framework for building student AI fluency grounded in Vygotsky’s zone of proximal development and the research on instructional scaffolding. It explains how AI can serve as a more knowledgeable other (MKO), but only when teachers set it up to scaffold: adjusting support to what the student already knows, fading help as the student grows, and handing responsibility back to the learner.
Drawing on recent studies of generative AI in classrooms, the article shows why default AI tools can boost short-term performance while weakening independent learning. It then walks through each STEP move, from AI modeling a task to the student proving they can do it alone, and maps each step to the ZPD and scaffolding research. Six original graphics illustrate the key ideas, including a two-track comparison of an AI that scaffolds and one that simply does the work.
The appendix includes a ready-to-use prompt that teachers can paste into a Google Gem, custom GPT, or similar tool to create their own scaffolding AI assistant. Written for educators and instructional leaders from K-12 through higher education, it pairs current research with practical guidance.
Read STEP: A Framework for Student AI Fluency (opens in new tab)