As a professor who researches the role of artificial intelligence in education, I’ve spent the past few years asking hard questions about how AI will shape teaching and learning. Will students learn less if they rely too heavily on generative tools like ChatGPT or Claude? How do we maintain academic integrity without becoming surveillance officers in our own classrooms? What does trust look like in an age of infinite assistance?
This past term, I had the opportunity to explore those questions not only as a researcher but also as a practitioner, in two very different business courses.
In a first-year Managerial Decisions course, a course focused on decision-making processes, my approach was simple: allow the use of generative AI, provide clear ethical and pedagogical boundaries, and focus less on policing and more on purpose.
At the start of the course, I told my students plainly, “Yes, you may use AI tools in this class under certain conditions.” We discussed how AI might help them brainstorm ideas, refine writing, or review concepts. But we also discussed the why of learning. I made it clear that the point of the course wasn’t to merely get through assignments; it was to build decision-making frameworks and evaluative thinking skills essential for business leadership. Tools may help with surface-level output, but if they bypass the hard mental work, they short-circuit growth.
I set expectations early and repeated them often. Throughout the term, I reminded students that while they had the flexibility to use AI responsibly, they were being prepared for a final exam in a fully controlled environment where no tools would be allowed. Whatever grades they might earn during the term by leaning on AI would become irrelevant if they couldn’t demonstrate understanding when it truly counted.
This pedagogical gamble was rooted in a hypothesis: if students rely too heavily on AI without developing real competence, their final exam scores will reveal the disconnect. However, if, even with access to AI, they internalize the learning goals and apply themselves with integrity, they will thrive because they’ve done the work and learning some of the desired skills.
What happened? The final exam came, and students sat individually in a proctored room with no devices, no AI, no shortcuts, just themselves and their minds.
And the class average? It was almost identical to the average throughout the term.
This result told me something profound: students had learned what they were meant to learn. Despite AI’s presence in the classroom, they showed real, lasting understanding of the material. Yes, some likely used AI more heavily than others. But the final assessment, designed to test applied knowledge and decision-making, filtered out the fluff. In the end, the students stood on their own, and most stood strong.
However, not every outcome was as consistent.
In the same term, I taught a fourth-year Data Analytics course with advanced students. I used the same approach: guided AI use, repeated expectations, clear ethical framing, and a final exam built to assess real understanding.
Throughout the term, the work was excellent, perhaps too excellent. I had several open and trust-based conversations with students and learned that several of them were using AI heavily. Again, I gave detailed examples of what to expect on the final exam, and I made it clear that their own learning, not AI’s assistance, would be evaluated.
This time, the results were different.
While the course average before the final exam was exceptionally high, the average on the final exam dropped significantly: one student failed outright. The exam was designed to assess real competence in data analysis, interpretation, and strategic thinking. Those who struggled were, not surprisingly, among the heaviest AI users.
There may have been several reasons. Some students, close to graduation, may have been focused only on maintaining their GPA or simply passing the course. Others may have relied too heavily on AI, inadvertently bypassing essential cognitive engagement. Even with clear expectations and sample questions, many were not able to demonstrate the depth of understanding that the course demanded.
This contrast does not disprove the value of integrating AI in learning. On the contrary, it confirms the core hypothesis: when students use AI to bypass learning, the consequences eventually appear. In this case, the final exam served as a reality check, revealing gaps in understanding that had been masked during the term.
These experiences do not ignore the challenges of academic integrity; they reframe them. Instead of treating students as adversaries, I approached them as adults, capable of moral reasoning and personal responsibility. Sometimes that trust resulted in meaningful engagement, and other times it did not, but both outcomes provided important insight.
In the Christian academic tradition, we are called not merely to transmit information, but to form character. That includes discernment in how we use tools, whether technological or otherwise. When we focus only on preventing dishonesty, we may win short-term battles but risk losing the deeper opportunity to build integrity and character. When we teach students why honesty matters and why it is worth struggling through difficult learning experiences, we affirm their agency and help support their long-term growth.
Ultimately, this reflects the example of Christ, who invites rather than coerces, and forms disciples through love, relationship, and responsibility. When we teach in a way that mirrors that model, we’re not just preparing students to succeed academically; we’re helping them grow in wisdom and character.
Good teaching is not about perfect control; it is about wise guidance. When we give students freedom with structure and autonomy with accountability, many will rise to the challenge. And when they do not, a carefully designed assessment can still reveal what is true.
AI is not going away, but neither is the power of thoughtful, faith-informed pedagogy. Let us continue trusting our students and showing them why it matters to be trustworthy. This is not a one-size-fits-all approach, but rather educational guidance rooted in character building and integrity. It is, in the end, a form of discipleship; teaching with patience, vision, and grace, just as Christ teaches us.
