For Prospective Families
AI Policy
Artificial Intelligence at Brunswick
A guide to how we think about AI — and why it matters
Artificial intelligence tools are now part of the world our students will inhabit, lead, and shape. At Brunswick, we approach this reality not with anxiety, but with the same values that have always guided us: Courage to engage honestly with hard questions, Honor to hold ourselves to the highest standards of integrity, and Truth to pursue genuine understanding rather than the appearance of it.
This guide is not primarily about rules. It is about what we believe. The policies that govern AI use at Brunswick exist because they reflect something deeper — a conviction about what education is for and what kind of people we are trying to become.
What We Believe About Learning
A Brunswick education is built on the relationship between a student and the hard work of understanding something. That struggle — sitting with a problem, revising an argument, rereading a passage until it yields its meaning — is not an obstacle to learning. It is learning. The moments when thinking is most difficult are often the moments when growth is most real.
AI can generate a polished essay. It cannot give you the ability to write one. AI can produce an answer. It cannot give you the judgment to evaluate one. The goal of education is not the product — it is the person who emerges from the process.
None of this means AI is unwelcome here. These tools are genuinely powerful, and students who understand them deeply will have real advantages in the world ahead. But that understanding requires critical engagement — knowing what AI actually does, where it fails, and when human judgment cannot be substituted. Those are lessons we can only teach if we are honest about what we value.
Our Five Principles
These principles apply to every member of the Brunswick community — students, faculty, and staff — whenever AI tools are part of our work.
1. Preserve the Human Work
AI supports thinking. It does not replace it.
The thinking, writing, analysis, and creativity that assignments are designed to develop must remain genuinely yours. AI can be a useful tool — for brainstorming, for feedback, for exploring a topic — but only when that use is authorized and disclosed. Submitting AI-generated work as your own is not a shortcut. It is a substitution of the very thing education is meant to build.
2. Act with Honor
Integrity means the same thing with AI that it always has.
When AI contributes to your work, you say so. You describe what it did and how you used it. This is not a technicality — it is the honest account of how your work came to be. Brunswick’s Academic Integrity Policy applies fully to AI use. The standard is not whether something can be detected. The standard is whether it is true.
3. Own Your Output
You are responsible for everything you submit, regardless of how it was produced.
AI tools can be wrong — confidently, fluently, and completely wrong. They can fabricate facts, invent citations, and reflect the biases embedded in their training data. If you submit work that contains AI-generated errors, the responsibility is yours. Verifying what AI produces is not optional. Neither is your own critical judgment about whether the output actually serves your argument.
4. Protect Privacy
What belongs to our community stays within it.
Personal information, student records, faculty communications, and Brunswick institutional data must never be entered into an AI platform. AI tools are cloud-based services — information submitted to them may be stored, used to train future models, or accessed by the platform provider. Our obligation to protect the privacy of every member of this community does not change because a tool is convenient.
5. Stay Vigilant
AI creates new risks as well as new possibilities.
Voice cloning, deepfakes, and AI-generated phishing are real threats that look increasingly convincing. Part of being an informed member of this community is knowing how these tools can be misused — and refusing to use them that way ourselves. Creating deceptive synthetic media, impersonating others, or using AI to undermine the trust our community depends on are serious violations of who we are.
Putting the Principles into Practice
Guidance for students, faculty, and staff
For Students
The most important thing to understand about AI at Brunswick is that context determines everything. The same tool can be a legitimate aid on one assignment and a violation of academic integrity on another. The determining factor is your instructor’s explicit guidance — and your own honesty about how you used it.
When AI may support your work:
- Brainstorming and generating initial ideas
- Getting unstuck on structure or organization
- Exploring a topic before forming your own view
- Receiving feedback on a draft you have already written
- Checking code logic or debugging
When AI is not appropriate:
- Any assignment where your instructor has not authorized it
- Writing, analysis, or arguments you submit as your own
- In-class assessments, quizzes, and examinations
- Work designed to demonstrate your individual understanding
- Any context where disclosure would not be possible
How to Disclose AI Use
When AI has contributed to your work and your instructor has authorized its use, your disclosure should be specific — not a vague acknowledgment, but an honest account. Include the tool used, what you asked it to do, and how you used or modified the output. The standard is this: could a reader understand exactly where AI ended and your own thinking began?
Example: “I used Claude to generate an initial list of counterarguments for my essay. I selected two that I found most compelling, researched them independently, and wrote my own response to each. All claims in this essay are verified by sources I consulted directly.”
For Faculty and Staff
The AI landscape asks something new of teachers: to think carefully about which parts of an assignment develop the capacities you most want students to build, and to be explicit about where AI fits — or doesn’t — in that work. Students cannot make good decisions about AI use if they don’t know what you expect. Clarity at the outset of every assignment is itself a form of teaching.
In your classroom
You are the authority on what AI use is appropriate for your assignments. That authorization must be explicit — silence is not permission, and students should never have to guess. Consider stating your position clearly at the top of each major assignment and revisiting it when expectations change.
Grading is always a human responsibility
- AI may inform your thinking, but the evaluation of student work rests entirely with you.
- No AI system may be used to determine grades, rank student performance, or make consequential academic judgments.
- This is not just a policy — it reflects what assessment is for: a faculty member’s professional knowledge of a student’s growth.
In your professional work
Faculty and staff may use AI tools for lesson planning, drafting communications, building rubrics, and administrative tasks — with the same standards of accuracy and integrity that apply to everything we do. The critical constraint: no student records, no personnel information, and no confidential institutional data may be entered into any external AI platform.
Be a critical user. AI output requires the same verification you would expect from any other source. If you would not cite a Wikipedia article without checking it, you should not rely on AI output without checking it. Your professional judgment is what makes AI useful — it is not what AI replaces.
Understanding What AI Actually Does
Much of the confusion around AI in schools comes from misunderstanding what these tools are. They are not search engines, and they are not experts. A large language model generates text by predicting what words are statistically likely to follow one another, based on patterns in its training data. It does not know whether what it says is true. It cannot verify facts. It does not have judgment.
This is why AI tools can be simultaneously impressive and unreliable. They produce fluent, confident prose about topics they have no actual knowledge of. They invent citations that look real. They can reflect biases present in their training data without flagging them. Every piece of AI output requires a human reader who knows enough to evaluate it — which is precisely the kind of knowledge a Brunswick education is designed to build.
The students who will use AI most effectively are not those who learned to prompt well.
They are those who built genuine expertise — and know the difference between AI that sounds right and an answer that is right.
This guide reflects Brunswick School’s Acceptable Use Policy for 2026–27 and is intended as a companion to it, not a replacement. Questions about specific AI tools or scenarios not addressed here should be directed to the Director of I.T. or the Head of School.
Brunswick School · Acceptable Use Policy Companion · 2026–2027
© Brunswick School – In Process of Revision and Review
