Designing AI-Resistant Assignments in Educational Leadership Courses – Faculty Focus

When I began teaching graduate-level courses in educational leadership early in the COVID-19 pandemic, my classroom existed entirely online. I quickly learned that keeping students engaged through a screen required more than well-organized slides or polished lectures. It required activities, learning experiences, and assignments that drew students into authentic, applied work that connected theory to the real leadership challenges they would face. Since then, I have taught dozens of educational leadership courses, both online and in person, and I have seen the same pattern: the design of an assignment shapes not only what students submit, but how deeply they engage with the learning process. Now, with generative AI able to produce essays, summaries, and even comprehensive education plans in seconds, the stakes for assignment design have changed. If an assignment can be completed entirely by an AI tool without any meaningful student input, we risk undermining the very skills that matter most in leadership preparation. 

Why AI-Resistant Assignments Matter

Some educators have begun using the term “AI-resistant assignments” to describe work that cannot be fully completed by a chatbot. These are not AI-proof, and I am not trying to exclude technology altogether. In fact, when used appropriately, AI can be a valuable support for brainstorming, organizing ideas, or refining language. But an AI-resistant assignment is one where the core of the work, including the evidence gathered, the decisions made, and the reflections offered, must come from the student’s own engagement with people, data, and context. 

Other educators have reached similar conclusions. In a recent New York Times column, Grose (2025) profiled professors who are reimagining their courses to ensure students remain active participants in their own learning. The examples she shared came from the humanities, but the underlying lesson applies across fields: in an AI-integrated world, we have to create assignments that cannot be outsourced entirely to a machine. The University of Massachusetts Amherst Center for Teaching and Learning (n.d.) offers strategies for making assignments more resistant to AI completion, such as requiring site-specific context, process documentation, and collaborative components. 

Further underscoring the need for evidence, the AutomatED newsletter issued a public challenge to educators to submit assignments they believed were “AI-immune” and then tested them directly with AI tools (Clay, 2023). The results demonstrated that only by actively trialing assignments against AI, not just assuming immunity, can instructors be confident in their design choices. 

These realities make it clear that designing assignments for leadership preparation cannot be left to chance. Educators need intentional strategies that not only account for AI’s capabilities but also reinforce the kinds of professional thinking and application AI cannot replicate. The following five principles offer a practical foundation for developing assignments that engage students deeply, foster leadership competencies, and ensure authentic performance. 

Five Principles for AI-Resistant Assignments

From my own practice and from published guidance, I have developed five principles for AI-resistant assignments in leadership preparation: 

1. Ground in authentic context – Require site-specific or scenario-specific evidence, incorporate the policy and community realities of schools, and engage peers or stakeholders in the process. This can include interactive formats such as presentations (recorded or live), simulations, debates, classroom observations with low-inference notes, analysis of instructional practice, interviews with stakeholders, or structured peer feedback sessions. 

2. Connect theory to practice – Anchor assignments in leadership frameworks such as the Professional Standards for Educational Leaders, Lencioni’s Organizational Health, or Kotter’s change model, and always include both application and contextualized reflection. 

3. Make the learning process visible – Assess drafts and planning notes alongside final products, and encourage multimodal evidence such as charts, agendas, or observation notes. 

4. Clarify the role of AI – Set explicit expectations for when AI can be used and when it cannot replace original, site-based analysis. 

5. Center human presence (Saucier’s PEACE Framework) – Build assignments that highlight preparation, expertise, authenticity, care, and engagement — qualities that cannot be automated and remain central to effective teaching (Saucier, 2025).  I make sure that my instructions are clear, grounded in tasks they will use in the field, and meaningful for them to complete in their future leadership positions. 
    

Examples from Leadership Preparation

These principles translate into assignments that blend conceptual rigor with applied relevance: 

• Community Introduction Town Hall – District leader candidates prepare and deliver a live presentation introducing themselves to the school community during an initial town hall meeting. The presentation includes school demographics, achievement data, community assets, and identified needs, and is followed by a brief Q&A. This assignment is explicitly aligned to the Professional Standards for Educational Leaders (PSEL) and requires candidates to connect their leadership vision to stakeholder priorities. 

• Organizational Health Rubric and Evaluation – Students use research to build on Patrick Lencioni’s organizational health model to create a custom rubric tailored to the context of a school district. They then apply the rubric to conduct an assessment, gathering both qualitative and quantitative evidence, rating each dimension, and concluding with a prioritized set of recommendations for improvement. 

• Clinical Internship Logs with Contextual Analysis and Showcase – Students maintain weekly logs documenting activities, decisions, and reflections from their clinical internship, ensuring each entry explicitly connects to course content and relevant standards such as the Professional Standards for Educational Leaders (PSEL). At the end of the course, they present a synthesis of their experiences in a live or recorded showcase, highlighting key takeaways, challenges addressed, and leadership growth. 

• Practice Instructional Coaching Conversations – Prepare for and conduct a simulated coaching conversation using an observation or case scenario, applying a framework like Jim Knight’s Impact Cycle, followed by reflection on the exchange. 
   

Implementation Strategies

When implementing these assignments, I sequence them so that simpler applications build into more complex, multi-stakeholder tasks, and I scaffold the skills of data collection, context analysis, and reflective practice. I apply UMass CTL’s transparency recommendations, making the purpose of each assignment explicit, clarifying AI use policies, and linking tasks directly to learning outcomes. I also keep the PEACE framework in mind in both my teaching and my assessment design. 

Conclusion: Keeping Humans at the Center

AI-resistant assignment design not only maintains academic integrity; it prepares leadership candidates for the realities of their profession. They practice making decisions with incomplete information, balancing competing interests, and communicating effectively with stakeholders, all in contexts where AI can be a partner but never the whole answer. By aligning these assignments with standards such as the Professional Standards for Educational Leaders and the National Educational Leadership Preparation standards, we ensure graduates leave with the judgment, adaptability, and human skills their roles demand. In an age when AI can produce passable work in seconds, including comprehensive education plans, the temptation for both students and instructors is to let the machine take the lead. But leadership is not about producing text; it is about guiding people and systems toward a shared vision. As Grose (2025) observed, the most enduring learning happens when students must engage with real people and real problems. If we want to prepare leaders who can thrive, we must ensure that every leadership course includes at least one AI-resistant assignment, not as a constraint, but as a commitment to developing the human capacities that no algorithm can replace. 

Andy Szeto, EdD, is an adjunct professor in educational leadership with experience teaching more than fifty graduate-level courses. He specializes in instructional leadership, school management, and AI integration in school leadership preparation, drawing from his professional background as a district leader and his ongoing work mentoring emerging school leaders.

References 

Clay, G. (2023, April 3). Believe your assignment is AI-immune? Let’s put it to the test. AutomatED: Teaching Better with Tech. 

Grose, J. (2025, August 6). These college professors will not bow down to A.I. The New York Times.

Saucier, L. (2025, May 12). What can college instructors offer their students in the age of AI? Faculty Focus. 

University of Massachusetts Amherst Center for Teaching and Learning. (n.d.). How do I redesign assignments and assessments for an AI-impacted world? Retrieved August 9, 2025, from