How the district defines academic integrity in an AI era, communicates it to staff and students, and operationalizes it in daily classroom work — covering both the binding layer (board-adopted definitions of authorized vs unauthorized AI use, attribution expectations, consequences, appeals process) and the practice layer (assignment design, fair-process protocols, classroom-level handling of AI assistance, the explicit recognition that AI-detection tools are unreliable evidence). Both dimensions are required: policy without practice creates a discipline minefield; practice without policy has no authority.
Why it matters
Legacy academic integrity language was written before generative AI. "Plagiarism" and "cheating" no longer map cleanly onto a landscape where AI can draft, edit, or co-author student work. Updated policy alongside aligned classroom practice keeps enforcement clear and fair, especially for students still learning the new expectations.
Connects to
The Framework: Cognitive & Ethical Foundation — Ethical Reasoning Under Real Pressure; Condition #3 (Mentoring & Modeling). Links to 1.3 (AI Policy) for broader governance scope. Distinct from 3.8 (Incident Response Protocols), which handles harm-related tech-mediated incidents (cyberbullying, deepfakes, sextortion) — academic integrity is a different domain entirely.
Maturity levels
Not Started
Policy language predates generative AI. No distinction between AI use and traditional plagiarism. No shared classroom practice. Teachers either ignore AI entirely or rely on unreliable detection tools. Enforcement is arbitrary across teachers and schools.
Emerging
Some policy language updated, often reactively. AI addressed as another form of cheating without nuance. Some teachers independently adapting assignments. No district guidance, no PD on AI-resilient design. Practice varies wildly across subjects and schools. AI-detection tools may be in use without explicit recognition of their unreliability.
Established
Board-adopted policy explicitly addresses AI assistance — defines categories of authorized use (brainstorming, drafting, editing, full generation), attribution expectations, consequences, and appeals process — differentiated by assignment type and grade level. Aligned classroom practice in place: shared guidance on AI-resilient assignment design, process-based assessment, fair-process protocols for suspected violations. PD on assignment redesign provided. Detection tools (if used at all) treated as one weak signal, never as proof. Policy and practice are visibly paired in handbooks, PD materials, and student-facing communication.
Expanding
Policy distinguishes categories of AI use precisely. Explicitly recognizes that prohibition without instruction is unenforceable. Practice has shifted from AI-detection to AI-transparent: students explain and reflect on their AI use as part of the assignment itself. Oral defenses, in-class drafting, process portfolios, and authentic performance tasks are common. Teachers share strong exemplars across subjects. Reviewed annually with student input. Public-facing plain-language summary available to families.
Go deeper with
Example resource
International Center for Academic Integrity (academicintegrity.org)
Also consider
MLA Guidelines for AI in Writing — practical teacher-facing guidance on assignment design and attribution
Common Sense Education AI Toolkit — teacher assignment-design resources
AI for Education — classroom-ready assignment templates and prompt libraries
Understanding by Design (Wiggins & McTighe) — backward design for AI-resilient assessment
Note: AI-detection tools (Turnitin AI, GPTZero, etc.) are unreliable and should not be the basis for disciplinary action