Zero Trust in the Classroom
I don’t want to be the technology police. As a teacher, policing and policies feel hollow. I may be in denial, but I don’t think they should be at the forefront of Christian education. I cannot see how using AI relates to the task of forming immortal souls.
As an educator, I don’t want to devote time to learning about and using large language models, but I also can’t ignore them. Within the last five years, every school district, every private school, and every home school co-op has revised its academic integrity policies with common phrases like “using artificial intelligence responsibly.” But what does artificial intelligence have to do with true education? And who is acting “responsibly” in this use—the teacher or the student?
Authentication
In computer science, two-factor authentication is a component of a security concept known as “zero trust.” It is a negative term because two-factor authentication is designed to protect accounts. That is what happens when a website or app sends you a two-factor authentication request, requiring a six-digit code every time you log in. It must ensure that you are who you say you are. The person trying to log in must (1) know something that only the account holder can know and (2) have something in their possession that only the account holder will have, be it phone, computer, or security key. This authentication is between the human and the system.
The zero-trust model in computer science, of course, uses basic facts. It is designed to protect accounts and information. It assumes an attack. At the same time, it is an authentication approach built on trust and knowledge, a concept mirrored in the student-teacher relationship.
As educators, we can ask if we trust the information a student gives us, but I think we must first ask if we trust the student. Formation in education cannot happen without it. The password can be stolen, but so can the second factor. Do I trust that my student wrote the words, created the spreadsheet, or produced the illustration?
The zero-trust model was originally designed, however, so that the second factor would be a physical object in the real, not virtual, world, something the owner possesses. A common example of this in the classroom would be the use of a blue book. I can clearly prove that a composition written in class is the student’s own thoughts and expressions since I am able to watch her write it in person.
In its security context, the zero-trust model means we must first verify before trusting. But how does the metaphor extend to the classroom? How do I verify? As a teacher, if I must prove something, that is, require proof, then I am not wholly able to trust. Here I am, in charge of young hearts made in the image of God, for a few hours a day, yet I might only see them through this disjointed lens—humans in a perpetual fallen state, unredeemed.
That type of mindset infects everything in my classroom environment as well as myself. If I first view every student as incapable of integrity, then something is wrong with me. Seasoned teachers are all too familiar with this landscape. So many of our students over the years have proved themselves untrustworthy that we educators become callous, full of mistrust. We become like the computer security systems that assume attack because of the proliferation of hacking and phishing. We come into the classroom assuming the worst, and the prevalence of ChatGPT has only made us more suspicious. That bias might be an unconscious weapon, but it still implies warfare.
The Arms Race
Alan Jacobs wrote about this same trust issue in a 2024 essay titled “Chatbots and the Problems of Life: Resisting the Pedagogy of the Gaps.” Teachers must now try to outwit their students and create assignments that make it hard to cheat. He terms it “an arms race,” where “[s]tudents use Internet technologies to enable teaching, and teachers call in other Internet technologies to detect that cheating.” Thus, trust becomes “a rapidly vanishing commodity.” The focus of class preparation becomes defensive. Instead of focusing on building relationships or on the quality of my content, I might have to plan how to counterattack when the learning process is abused. In recent years, that same negative language already has become ubiquitous. In her essay “A Matter of Words: What Can University AI Committees Actually Do?”, Megan Fritt assumes a combative tone when she describes university AI policies: “[O]ne first has to have an idea of which skills and formative experiences they are prepared to lose for the sake of AI use, and which ones they will fight to retain.”
In the greatest irony, college students have accused teachers of inaccurately using AI to detect AI. At the website Inside Higher Ed, articles like “Professors Cautious of Tools to Detect AI-Generated Writing” mention multiple instances at universities that are simply causing more confusion. At Harvard’s metaLAB AI Pedagogy Project, the July 2023 Sample Code of Conduct says such use “can both lead to missed instances of AI-generated text and unfairly implicate students who did not use AI in their writing. Furthermore, trying to ‘detect’ academic dishonesty in this way conveys a lack of trust of students in an academic environment that is already based on an Honor Code.” Similar articles and statements across American university campuses abound.
Trust and Learning
I, and many other educators, would like to see teachers and students creating a working shared key encryption to guarantee that a student wrote and completed the assignment that was turned in, but truly I don’t want to have to outwit my students or play offensive war games. I want to help them learn. I want them to know that education isn’t about grades and jumping through the hoops to get to the next stage of their education. They are trying to get through something instead of truly learning. That view of learning leaves little room for creativity, wonder, and intellectual growth.
I want to build an environment where we trust each other first. Like Ted Gioia, as explained in his essay “5 Ways to Stop AI Cheating,“ I will continue to use “old-school” assignments in my Humanities classes—oral quizzes, Oxford seminars, graded discussions, and in-class writing exercises. I may never experiment with Chat in the classroom because I value my class time too much. That doesn’t mean I won’t address academic integrity. That will always be part of any classroom. But not every teacher has the freedom to choose whether they use or model artificial intelligence. Many schools require its use in some way, whether for grammar checks, math help, or basic programming. In an article titled “Being Human in the Age of AI,” Jen Pollock Michel aptly writes, “Even if we might forfeit the use of a tool for ourselves, we can’t reverse the expectations it creates in society at large. Whether we embrace the efficiencies of text generators for our own tasks, we aren’t likely to escape the appetite for quick fixes of other kinds.”
In the Christian school or university classroom, I see one way out, or better said, one way in. Zero trust is not a zero-sum game. In game theory, it is closer to the Nash Equilibrium where players have chosen their strategies. No player can benefit by changing strategies while the other players’ strategies remain unchanged. It’s one way of looking at working together. Trust or mistrust doesn’t simply happen to us. It is a choice. In the classroom, someone has to change and lead the charge, and it must be the teacher. It is a different geography. It is not a student-led classroom but a teacher-led one. It is not a broad worldview but a honed view of the classroom world.
The classroom world is one of intentional design. Effective teachers know this and choose what does or doesn’t happen in that space. How they speak, how they allow students to speak, how they prepare for their lessons, and how they present them are all done with intention.
Intention breeds trust because it shows our care for those we are responsible for teaching, guiding, and shepherding. Our personal and spiritual values as teachers, along with shared experiences with our students, communicate to them what we believe is most important about the curriculum we teach, who we are, and the God we worship.
Whatever we may think of the academic integrity of the student, I must look at my own integrity first if I am to continue to build trust. Instead of suspicion and accusation, my classroom must be built on love. This means that relationships come first. In a classroom world where love gives and receives, if I speak harshly or out of frustration, I can ask my class for forgiveness. If I communicate the wrong directions or forget to explain something fully, I humble myself and amend the mistake. I hope my students can do the same.
I think of the words of Augustine, Clement of Alexandria, John Chrysostom, and others amid their modern echoes in books like Alex Sosler’s Learning to Love: Christian College as Pilgrimage and David Smith’s Everyday Christian Teaching: A Guide to Practicing Faith in the Classroom. Doubt, suspicion, and anxiety must not outweigh love in the classroom. Zero trust has no place, but authentic relationships do. For what we love most or fear most influences the life we pursue, inside, and beyond, the classroom.