Before a single student logged on this academic year, the Diocese of Arlington had already drawn a line in the sand. A working group released a formal policy - "Responsible Use Policy for Technology and Artificial Intelligence" - ahead of the 2025-26 school year, making Arlington one of the first Catholic dioceses in the United States to set institutional rules governing how artificial intelligence may be used across its schools. The policy attempts something genuinely difficult: giving students access to one of the most consequential technologies of their generation while ensuring that technology serves human development rather than shortcutting it.
Defining the Purpose: AI as Tool, Not Teacher
The central premise of the diocesan policy is carefully worded. Artificial intelligence is to be treated as a tool and a "thought partner" - a phrase used by Leslie Lipovski, diocesan assistant superintendent of student learning and teaching excellence. The distinction matters. A tool augments human capability; it does not replace human judgment. And in a Catholic educational context, that distinction carries additional weight, connecting to broader questions about human dignity, moral agency, and what education is actually for.
At the elementary level, the approach is deliberately constrained. The Children's Online Privacy Protection Act, a federal law passed in 1998, restricts the collection of personal data from children under 13, which limits the AI platforms schools can deploy for younger students in the first place. The diocese has used that legal reality as a philosophical anchor: for kindergarten through eighth grade, the goal is AI literacy - understanding what the technology is, how it works, where it succeeds, and where it fails - rather than direct, hands-on use of AI tools. Teachers may use AI to adapt lesson materials, adjusting reading levels for individual students or generating activity ideas, but students are not logging into chatbots themselves.
The policy also states explicitly that technology will never replace teachers, a commitment that sounds obvious until one considers how many institutions have quietly allowed algorithmic systems to absorb functions once performed by educators. The policy will be updated as the technology changes - a necessary clause given how rapidly the landscape shifts.
High Schools Test the Boundaries in the Classroom
At the secondary level, the approach becomes more direct. St. Paul VI Catholic High School in Chantilly has developed a program called LUMA that integrates with the Google Classroom platform, tested this year by roughly 20 students enrolled in an Introduction to AI course. Saint John Paul the Great Catholic High School in Potomac Shores is preparing to offer a dual enrollment bioethics course titled "AI and Algorethics," which will ask students to examine the ethical stakes of algorithmic systems - questions of human dignity, privacy, surveillance, and what the course describes as the transparency of algorithms.
That last concern is not abstract. Algorithmic opacity - the tendency of AI systems to produce outputs without explaining their reasoning - is one of the most consequential issues in applied AI today, affecting everything from medical diagnosis to criminal sentencing. Teaching high school students to interrogate that opacity, rather than simply accept it, reflects a more ambitious pedagogical ambition than most institutions have managed.
"While we are embracing what AI may have to offer, we want to ensure it is used to serve humanity and not diminish it," said Greg Haas, assistant principal for academics at Saint John Paul the Great. That framing - AI in the service of the authentic person - runs throughout the diocesan approach and distinguishes it from the more purely technical or productivity-oriented frameworks adopted by many public school systems.
The Classroom Problem: Plagiarism, Shortcuts, and the Integrity Question
The policy's aspirations meet their hardest test in English and humanities classrooms, where the line between AI-assisted thinking and AI-substituted thinking is frequently invisible. Rebecca Vaccaro, head of the English department at Bishop Ireton High School in Alexandria, has spent three years watching student reliance on AI increase. Students submit paragraphs generated by chatbots, sometimes incorporating AI-written material into papers that are otherwise their own, sometimes submitting work that is almost entirely machine-produced. Detection is unreliable. The writing patterns that once flagged plagiarism have shifted.
Vaccaro's response has been structural: handwritten in-class assignments, low-stakes "discovery writing" that prioritizes process over product, and explicit classroom discussion of what ethical AI use actually looks like. History teacher Garrett Fabacher at the same school has developed assignments specifically designed to be "AI-resistant" - tasks requiring personal observation, contextual reasoning, or original synthesis that a language model cannot convincingly replicate.
At St. Veronica School in Chantilly, Latin teacher Jenifer Scott and literature teacher Kandis Rouck have returned to composition books and pens. The choice is deliberate. Writing by hand, they argue, protects the cognitive process that makes writing worthwhile. Rouck put it plainly: reaching for an AI tool interrupts the mental flow that produces genuine thought. Scott made the same point about Latin translation - using an AI-powered translation service produces a correct answer and teaches nothing. The value of the exercise is in the struggle toward the answer, not the answer itself.
What This Policy Signals for Institutions Beyond the Diocese
The Arlington diocese's approach is notable less for any single decision than for the coherence of its underlying framework. Most institutional AI policies written in the last two years have been reactive - produced in response to a specific incident or external pressure, focused on prohibition rather than pedagogy. This one begins from a stated set of values and asks how AI fits within them, rather than asking how to stop students from misusing a tool that has already arrived.
Second grade teacher Jonathan deBernardo from Queen of Apostles School in Alexandria offered a pointed example of what that value-driven caution looks like in practice: a chatbot asked to simplify a scripture passage for young readers may return a version that distorts the original meaning or, in his word, is outright heretical. The prompt shapes the output, and the educator must understand both well enough to catch the difference. That is not a technical skill. It is a critical one - and the kind that AI literacy, properly taught, is designed to build.
The broader lesson from Arlington may be that institutions which begin with clear values have a more stable foundation for AI policy than those chasing the technology itself. As the tools change - and they will change significantly - the questions of human dignity, authentic learning, and ethical use will remain constant. That is, perhaps, exactly what a liberal arts education grounded in moral reasoning is designed to prepare students to answer.
