AI Principles and Policies
The following text defines the principles and policies for the use of AI in the School of Architecture and is intended as a reference for instructors, students, and administrators. It is comprised of three parts: Principles (based on the work of the AI in the Arts Working Group) that outlines broad concepts about AI use; Policies, defining specific protocols and rules for AI use and mis-use in the School; and Model Syllabus language, proposing language to be included in each course syllabus.
The working group recommends that Introduction, and Principles and Policy sections be included in the Handbook, the Introduction and the Policies included in the Bulletin, and that this entire document, including recommended syllabus language, be distributed to the entire faculty.
Introduction
The use of artificial intelligence at the School of Architecture is governed by the following principles and policies. It applies to all students, faculty, and staff in the preparation of academic, design, and research work.
Principles
Artificial intelligence comprises a range of technologies with wide potential use across the School of Architecture. As with any significant trend or transformation in our discipline, the School of Architecture is committed to understanding, exploring, and defining the theoretical and practical implications of AI in architecture. We therefore encourage the ethical use of these tools in our pedagogy and research, based on the following principles:
AI as Tool, Not Author: AI may serve as a resource for ideation, research, iteration, and exploration, but the artistic intent, creative voice, and intellectual labor at the center of the work must remain human. Students, faculty, and staff are expected to author their own processes and articulate how and why AI contributed to any given work.
Transparency and Attribution Are Non-Negotiable: The substantive use of AI in the preparation of coursework, research, performance, or creative work must be acknowledged. This applies regardless of how substantially the AI-generated content was altered or integrated. Undisclosed use constitutes a violation of academic integrity.
Course Guidelines Define the Terms of Engagement: Faculty set AI policy for their courses, from open experimentation to explicit prohibition, in alignment with their pedagogical goals and disciplinary values, and students are responsible for understanding and honoring those terms. University guidance, including resources from the Poorvu Center for Teaching and Learning, provides a shared foundation.
Critical Engagement Over Passive Use: Yale Architecture, Art, Drama, and Music must prepare artists, scholars, and practitioners for a world in which AI is already present. That preparation includes developing the judgment to use AI thoughtfully, question its outputs, understand its limitations and biases, and resist the flattening of creative work toward the generic, easy, or convenient.
Human Development Is the Purpose: AI should support learning, not shortcut it. The goal of graduate education in the arts is the cultivation of deep listening, interpretive judgment, embodied practice, original thought, and collaborative capacity. Any use of AI should be evaluated against whether it advances or undermines that development.
AI is changing the Culture and Practice of Architecture: As with any significant trend or transformation in our discipline, the School of Architecture is committed to understanding, exploring, and defining the theoretical and practical implications of AI in architecture. Ai is an external force that should be engaged with both criticality and innovation.
Policies
The following policies govern the use of artificial intelligence in the School of Architecture:
Proper Use of AI Tools
Unlike previous digital tools, AI instruments, ranging from large language models to generative algorithms to autonomous agents, may act autonomously and exercise authorial capabilities. For this reason, any use of AI that materially contributes to the substance of submitted work must be acknowledged and, where required, documented, in accordance with the University and School of Architecture academic integrity, plagiarism, and attribution requirements, see https://www.architecture.yale.edu/academics/school-handbook “Unacceptable Conduct.” Improper use of AI will be treated as a violation of academic integrity, which may include plagiarism, unauthorized assistance, or fabrication, and is subject to disciplinary action.
Course Policy Requirements
Instructors must stipulate the AI policy for their specific course in their syllabus and explain it at the beginning of the term. These terms may range from explicit prohibition to broad permission. Irrespective of course policy, students must acknowledge any use of artificial intelligence in the preparation of their work and, where required, provide specific documentation, which might include, for example, prompts, AI-revised text, AI-generated code, or AI-generated images, renderings, and design output. Instructors may identify routine, embedded uses (such as basic spelling and grammar correction) that do not require acknowledgment. Faculty are directed to resources at the Poorvu Center for Teaching and Learning for further guidance.
Disciplinary Actions
Violations of this AI use policy will be referred to the Rules Committee and are subject to the disciplinary rules and procedures of the School.
Model Syllabus Language
Instructors are required to state an AI policy for their course. The following options may be adopted directly or adapted. Each option inherits the School-wide requirement that AI use be acknowledged; even where AI use is prohibited, undisclosed use is itself a violation. Instructors who wish to exempt routine, embedded uses (see below) should append the non-attribution note to their chosen option.
Option 1: Use Prohibited
Artificial Intelligence. The use of artificial intelligence tools in any work submitted for this course is prohibited. All work, including text, code, drawings, renderings, and design output, must be your own. You may not use large language models, generative image or design tools, or AI agents to produce, revise, or resolve
any portion of your submitted work. If you are uncertain whether a particular tool falls under this prohibition, ask before using it. Consistent with School of Architecture policy, any use of AI must still be disclosed; undisclosed use, or use in violation of this policy, will be treated as a violation of academic integrity and referred to the Rules Committee.
Option 2: Refinement of Student-Originated Work
Artificial Intelligence. In this course, artificial intelligence may be used only to refine, develop, or resolve work that you have originated. The conceptual substance, including the argument, the design, the analysis, must originate with you; AI tools may assist in editing, clarifying, debugging, or otherwise improving work you have already produced. AI may not be used to generate primary content, originate design concepts, or substitute for your own thinking. You must acknowledge any use of AI in your submission and, when asked, document it by providing the relevant prompts, the AI-revised text or code, or the original work that preceded the AI-assisted revision. Using AI to originate rather than refine work, or failing to disclose AI use, will be treated as a violation of academic integrity and referred to the Rules Committee.
Option 3: Permitted with Attribution
Artificial Intelligence. In this course, artificial intelligence tools may be used freely, provided that all such use is properly attributed and documented. You remain fully responsible for the accuracy, quality, and integrity of everything you submit, including any errors, fabricated citations, or design flaws introduced by an AI tool. For each submission, include a brief statement describing which tools you used and how, such as the tasks delegated to AI and a representative sample of prompts. When asked, you must provide fuller documentation, such as relevant prompt histories, AI-generated text, code, images, or design output. Undisclosed or undocumented use will be treated as a violation of academic integrity and referred to the Rules Committee.
Model Acknowledgment Statement
Instructors adopting Option 2 or Option 3 should direct students to use a standard disclosure format so that acknowledgment is consistent across courses:
“I used [tool] to [task]. Representative prompts and outputs are included / available on request.”
Optional Non-Attribution Note
Instructors may append the following to any option to exempt routine, embedded uses from the acknowledgment requirement:
Routine, embedded features, such as spelling and grammar correction, autocomplete, search, and the AI capabilities built into standard design software (e.g., Rhino, Revit, Adobe applications), do not require acknowledgment under this policy.
A Note on Studios and Seminars
Instructors should consider the medium of the course when selecting an option. A prohibition on AI text generation in a history or theory seminar is straightforward. In studio courses, generative image and design tools are increasingly part of the design conversation itself, and “prohibition” or “free use” may carry different meaning than in writing-based courses. Instructors must Tbe explicit about how their chosen policy applies to design output specifically.