AI Use Tiers

Goal

Part 2 of the AI Governance Guidelines: a shared vocabulary for how much AI is allowed on any given assignment.

The AI Use Tiers give the institution a single shared vocabulary for the question students ask most: "Am I allowed to use AI on this?" Rather than every course inventing its own rules, each assignment is tagged with one of five tiers. The tier travels with the assignment, named in the syllabus, repeated in the assignment prompt, and reflected in the rubric.

How to apply: an instructor selects a tier per assignment (not per course). A single course will often span several tiers. For example, Tier 0 for an in-class exam and Tier 3 for a project. When no tier is stated, Tier 1 is the institutional default.

2.1 The five tiers at a glance

Tier	Name	What students may do	Disclosure required
0	No AI	Work entirely independently. No AI assistance of any kind (e.g. in-class exams, foundational skills checks).	N/A — AI use is a violation.
1	AI for Ideation	Use AI only to brainstorm, explore topics, or plan. All researched, written, and submitted content is the student's own.	Brief note of how AI was used for ideation/planning.
2	AI-Assisted	Use AI to support the work — outlining, feedback on drafts, light editing. Substance and final expression remain the student's.	Statement of how and where AI was used; instructors may request prompts/outputs.
3	AI Collaboration	Co-create substantial portions with AI (e.g., drafting, code, analysis) while directing, verifying, and revising. Process is part of the grade.	Full process record: prompts, key outputs, and a reflection on edits/decisions.
4	AI-Led / Open	Use AI freely, including AI-generated final output, as the explicit point of the task (e.g., evaluating model behavior, AI-fluency exercises).	Disclose tools and approach; cite AI as a source when applicable.

Color cue: red = restricted, amber = supported with limits, green = collaborative/open. Use the same colors in assignments so the tier is recognizable at a glance.

2.2 Tier definitions in detail

Tier 0 — No AI

AI tools may not be used in any part of the task. Reserved for assessments measuring foundational, independent capability: closed-book exams, in-class writing, skills demonstrations, and diagnostic work. Use of AI here is an academic-integrity violation. Enforce Tier 0 through secure conditions (proctoring, in-class, or handwritten work), not through AI-detection software, which is not an approved assessment mechanism (see AI Policy for Students).

Tier 1 — AI for Ideation

Students may use AI to get unstuck, i.e. generating topic ideas, exploring angles, or structuring a plan, but the research, writing, and final product must be entirely their own. This is the institutional default when no tier is specified, because it preserves independent work while acknowledging AI as a legitimate thinking partner at the planning stage.

Tier 2 — AI-Assisted

AI supports the student's own work: drafting feedback, summarizing source material the student then verifies, or light editing. The intellectual substance and final expression must remain the student's. This tier covers most everyday coursework where AI improves quality and efficiency without doing the thinking. Disclosure is expected, and instructors may ask to see prompts and outputs.

Tier 3 — AI Collaboration

Students actively co-create with AI, i.e. generating drafts, code, data analysis, or design that they then direct, critique, verify, and substantially revise. Because the value lies in the collaboration, the process itself is assessed: students submit prompts, key AI outputs, and a short reflection explaining their decisions and edits. This is the home of the most sophisticated AI-integrated assignments (simulations, custom bots, iterative prompt exercises).

Tier 4 — AI-Led / Open

Unrestricted AI use, including AI-generated final output, where engaging deeply with AI is the explicit learning objective. For example, evaluating model strengths and weaknesses, AI-fluency labs, or producing an artifact to critique. Students disclose their tools and approach and cite AI as a source. Even here, students remain accountable for accuracy, ethics, and the claims they make.

2.3 Disclosure norms

Disclosure is the connective tissue of the whole framework. Acceptable methods, scaled to tier, include:

A short AI-use statement appended to the assignment ("I used [tool] to …").
In-text citation of AI per the relevant style guide (e.g., MLA/APA).
Screenshots or links of prompts/chats.
A reflective log or logbook documenting AI interactions and reflections throughout a project (recommended for Tier 3).
Sharing the working chat when the approved platform supports it (i.e. BoodleBox).
Including reflection on AI's output, human modification, and why AI was applied.

Questions?

Email compliance@boodle.ai for governance specifics, or success@boodle.ai — typical response within 8 business hours.