James D’Annibale, Director of Academic & Emerging Technologies

Originally written July 2025. This version was updated July 2026. The changes to the guidance portion are minimal and reflect changes to available technology and changes in higher education pedagogical practices. The “Example Assignment Briefs” have been totally overhauled to reflect real assignments done at Dickinson College over the last year.

Purpose of This Guidance

Pedagogical transparency helps ensure that students:

• Understand the learning objectives behind each major assignment.
• Recognize how completing the assignment contributes to long-term intellectual and professional development.
• Are aware of which aspects of the assignment must be completed independently, and which may involve generative AI or other forms of assistance, including assistance from student peers.

This guidance is intended first and foremost to support student learning. When students understand the goals of an assignment and how they are expected to engage with it, they are better equipped to succeed. Research on pedagogical transparency has shown that when students clearly understand the purpose, tasks, and evaluation criteria of assignments, they experience greater academic confidence, stronger learning outcomes, and improved metacognitive awareness (Winkelmes, M. (2013). Transparency in Teaching: Faculty Share Data and Improve Students’ Learning. Liberal Education 99(2).). In this context, we extend the concept of transparency to a contemporary challenge: helping students navigate the appropriate use of generative AI tools. Generative AI is increasingly integrated into many of the digital tools students already use for writing, research, communication, productivity, and creativity. The principles in this guidance apply regardless of which AI-enabled tool or platform a student chooses to use. By making expectations and rationales explicit, instructors may reduce the likelihood of inadvertent academic integrity violations in this evolving landscape.

The work of athletic coaches is instructive when thinking about how pedagogical transparency can help students learn and make choices regarding shortcuts. When a student-athlete works out in the gym, it’s very clear to the student-athlete what is expected of them and how hard work and meeting/exceeding expectations will benefit them as individuals and as a member of a team. Athletes and teams that work hard in the gym and on the practice field, often see great results on the game field/arena while the opposite is true for athletes and teams that do not work hard.

The same clarity is being encouraged in this guidance document. We ought not assume that our students understand the purpose of assignments within the context of their learning or within the context of the liberal arts. Rather, we should be clearly explaining purpose, benefits, expectations, etc. so that our students truly understand why we’ve drawn whatever line in the proverbial AI sand that we’ve chosen to draw.

Guidance for Transparent Assignment Design

Faculty designing major assignments should ensure the following elements are communicated to students:

• Learning Objectives: Clearly state, in understandable terms, the specific skills or knowledge the assignment is meant to develop.
• Connection to Long-Term Learning: Briefly explain how the assignment supports the student’s academic, professional, or personal growth.
• Student-Generated Work: Indicate which portions of the assignment must be completed independently and why.
  • In some cases, it may be helpful to guide students in breaking down the assignment into steps that align with learning objectives, course content, and intended skills.
• Permitted Use of AI or Other Help: Clarify what kinds of assistance (AI-enabled tools, tutoring, peer review, etc.) are acceptable, what role that assistance is intended to play in student learning, and in which stages it may be used. When AI is permitted, students should also understand what thinking, decision-making, and responsibility remain their own.
• Disclosure Requirement: If AI is permitted, explain how students should document or reflect on their use in a manner appropriate to the assignment (e.g., a Panopto video journal in which the student displays and talks through their work with AI, an appendix, an in-paper note or citation, a brief reflection, or another instructor-approved method).
• Invitation to Clarify: Encourage students to ask questions if they are unsure about the policy. Also, welcome students to share their own ideas for how they might use AI to enhance their learning, so that faculty can consider and respond to these suggestions constructively.
• Faculty Presence in the Process: Incorporate specific opportunities to engage with students throughout the development of the assignment (e.g., requiring outlines, holding progress check-ins, reviewing early drafts). This builds a learning relationship, supports student growth, and allows instructors to better understand the student’s thinking and voice as it evolves—helping evaluate the final product within the context of the process. The expert staff at the Writing Center and the Center for Teaching, Learning, and Scholarship are tremendous resources in this area.

Example Assignment Briefs

The examples below are intentionally brief and meant to serve as starting points. They illustrate different roles AI might play while preserving meaningful intellectual work and responsibility for students.

Example 1: Position Paper with a Skeptical Discussion Partner

Learning Objectives: Develop a position, evaluate competing perspectives, respond to objections, and revise an argument based on careful reasoning.

Long-Term Value: Prepares students to explain and defend their ideas, engage constructively with disagreement, and determine when criticism should influence their thinking.

AI Use Policy: After developing an initial position, students may use AI as a skeptical discussion partner. The AI may question assumptions, raise objections, request clarification, and challenge the reasoning behind the student’s position. It should not determine the student’s position or decide whether the argument is correct.

Disclosure: Students should briefly explain which objections or questions were most useful, how they evaluated them, and whether the interaction caused them to revise their argument.

Faculty Presence in the Process: Students submit an initial position or working thesis before beginning the AI interaction. The instructor reviews the student’s developing argument and may discuss how the student responded to the challenges the AI raised.

Example 2: Revising a Reflection After Class Discussion

Learning Objectives: Recall and synthesize a class discussion, connect the ideas of others to one’s own thinking, and explain how an initial perspective developed or changed.

Long-Term Value: Helps students become more thoughtful participants in collaborative learning and better able to recognize how conversation, evidence, and differing perspectives influence their thinking.

AI Use Policy: Students first write their own reflection and participate in the class discussion. They may then use an instructor-designed AI interaction that asks questions about what they contributed, what they heard from others, and how their thinking evolved. AI may help students organize the ideas they identify, but it should not invent details from the discussion or determine what the student learned.

Disclosure: Students should include a brief note describing how the questioning process helped them reconstruct or reconsider the discussion.

Faculty Presence in the Process: The instructor has access to the original reflection and establishes the questions or boundaries that guide the AI interaction. The revised reflection can therefore be considered in relation to the student’s original thinking and participation in the course discussion.

Example 3: Psychological Analysis of a Simulated Character

Learning Objectives: Apply psychological theories, develop meaningful interview questions, evaluate evidence, and recognize inconsistencies or limitations in a subject’s account.

Long-Term Value: Develops skills in interviewing, interpretation, theoretical application, and evidence-based analysis.

AI Use Policy: Students may interview an AI that has been instructed to role-play a fictional character based on events, personality, and history established in an assigned text, film, or television series. Students must develop their own questions, evaluate the character’s responses, and support their analysis with evidence from the original source. AI responses should be treated as material for analysis rather than authoritative evidence about the character.

Disclosure: Students should identify how the simulated interview informed their analysis and distinguish between evidence drawn from the original source and statements generated during the AI interaction.

Faculty Presence in the Process: The instructor establishes the parameters of the simulation and reviews students’ proposed questions or initial theoretical approach before the interview. The final analysis is evaluated based on the student’s application of theory and use of evidence, not on the apparent realism of the AI character.

Example 4: Scientific Manuscript with AI-Assisted Drafting and Review

Learning Objectives: Synthesize scientific literature, organize a manuscript, distinguish supported claims from overstatements, and take responsibility for scientific accuracy.

Long-Term Value: Develops skills in scientific writing, critical reading, source evaluation, and the responsible use of digital tools in research-based settings.

AI Use Policy: Students must first select and read the research articles, take meaningful notes, and develop their own ideas about how the manuscript should be organized. They may then provide those materials, the assignment instructions, and the source articles to AI to help create a detailed, source-mapped outline. After reviewing and revising the outline, students may use AI to generate a source-grounded first draft. AI may also be used as a scientific review partner to identify claims that require verification, possible overstatements, unsupported conclusions, or ignored limitations.

Students must independently verify every factual claim and citation against the original research. AI output is treated as raw material rather than finished work, and responsibility for the final scientific argument remains with the student.

Disclosure: Students should document the major stages of their process and explain what they accepted, revised, rejected, or corrected. A Panopto video journal may be especially useful for allowing students to display their work while explaining the decisions they made.

Faculty Presence in the Process: Students will meet with the instructor during the proposal, outline, and near-final stages to discuss how they are using AI, what challenges they’ve encountered, and how they’re ensuring scientific accuracy in their work.

Example 5: Evidence-Based Analysis of a Complex Problem

Learning Objectives: Develop hypotheses, identify relevant evidence, interpret data, recognize limitations, weigh competing priorities, and defend recommendations.

Long-Term Value: Prepares students to address complex problems for which there may be multiple explanations, incomplete evidence, and no single objectively correct solution.

AI Use Policy: Students must identify the problem, develop their own hypotheses, and decide which questions or analyses are relevant. AI may assist with technical analysis of a supplied dataset and may question students about whether the evidence supports their interpretations. It may identify patterns, probe their logic, request clarification, or point out where additional evidence may be needed. AI may not determine what the results mean or what actions should be recommended.

Disclosure: Students should explain how AI supported the technical analysis, how they evaluated its output, and which interpretations and recommendations originated from their own judgment.

Faculty Presence in the Process: The instructor reviews students’ initial hypotheses before the analysis and provides opportunities for students to present, defend, and revise their interpretations. The final work is evaluated based on the quality of the students’ reasoning and recommendations rather than the complexity of the technical analysis produced with AI assistance.

Example 6: Language Conversation Practice

Learning Objectives: Listen, speak, respond, and sustain a conversation using vocabulary, grammar, and cultural knowledge appropriate to the course.

Long-Term Value: Gives students repeated opportunities to develop fluency, confidence, comprehension, and the ability to respond spontaneously in the target language.

AI Use Policy: Students may engage in spoken or written conversations with an AI conversation partner configured by the instructor for a specific course or unit. The scenario should remain within the vocabulary, grammar, communicative skills, and cultural context students are expected to practice. AI provides an additional opportunity for conversation but does not replace interaction with the instructor or other students.

Disclosure: Students may be asked to submit a brief reflection identifying moments when communication was difficult, strategies they used to clarify meaning, and areas they would like to practice further.

Faculty Presence in the Process: The instructor establishes the scenario and its linguistic parameters, reviews patterns in student performance, and uses those observations to inform later instruction, feedback, or in-person conversation practice.

Introducing Transparency to Students

For faculty who are beginning to integrate greater transparency about assignment design and AI use, it’s important to build trust, clarity, and student engagement. Below are some suggested steps to help introduce these concepts, especially in courses where this level of openness may not have been present previously:

• Acknowledge the Shift: Let students know that you’re working to be more explicit about expectations and learning goals. A simple statement like, “I’ve revised how I present assignments to help you better understand not just what to do, but why it matters,” can be effective.
• Explain the Rationale: Be clear about why certain parts of the assignment must be done independently, why AI has been assigned a particular role or intentionally excluded, and where AI or other forms of help may be beneficial. Framing this around their long-term learning and success helps build buy-in.
• Invite Discussion: Consider hosting a short class discussion or informal survey about how students have encountered AI tools and where they might see value in using them. This gives you insight and models an open learning environment.
• Introduce Policies in Context: Rather than just listing AI rules, walk students through a specific assignment and explain which parts are meant to develop their own thinking, why that thinking matters, versus where AI could be used to support workflow or clarity.
• Model Transparency: Briefly explain your instructional design choices. For example: “I’ve designed this project so that you get practice developing your own argument. That’s why the initial analysis must be your own, but I’ll let you use AI to help polish the writing after you submit your draft.”
• Stay Open to Feedback: Let students know you’re learning too. Invite them to let you know what works and where things could be clearer.

By taking time to introduce your approach and show its connection to student learning, you help reduce confusion, prevent integrity issues, and support a healthier academic environment.

Link to AI Syllabus Statement Guidance

Faculty should ensure consistency between their assignment-specific guidance and the course-wide AI policy stated in their syllabus. Sample language for syllabus policies can be found in the Sample AI Syllabus Statements document.

Written by James D’Annibale with assistance from ChatGPT. AI served as a discussion partner throughout the writing process, helping critique ideas, identify places where current technology and pedagogy suggested updates, and improve the clarity and organization of the final document. All recommendations were reviewed by the author, who remained responsible for the content of the final guidance.