Skip to Main Content

For Faculty

AI in Assessment

This guide is designed to support Emerson College faculty in designing thoughtful, transparent, and outcomes-aligned assessments in the age of generative AI. As artificial intelligence tools like ChatGPT and DALL·E become more accessible to students, educators are rethinking what it means to demonstrate learning—and how to assess it.

What You'll Find Here:

  • A clear explanation of Furze et al's AI Assessment Scale

  • Guidance on mapping AI use to your course learning outcomes

  • Tools for building AI-responsive rubrics

  • Resources for further exploration and support

Why Use This Scale?
Leon Furze’s scale helps educators intentionally position AI use in their courses—ranging from AI-Free to AI-Generated—depending on the purpose of the assignment, discipline norms, and pedagogical goals. This guide helps you use the scale as a flexible tool, not a prescription, to deepen student learning.

Overview of the AI Assessment Scale

This five-level framework offers a flexible way to position generative AI tools within your assignment design:

Level Description
(1) AI-Free No AI tools allowed; emphasis on original human though and process.
(2) AI-Aware Students must acknowledge or cite any use of AI (e.g., proofreading, idea generation).
(3) AI-Optional Students may choose to use AI as a supplemental tool; guidance is provided.
(4) AI-Integrated AI use is expected and purposefully built into the assignment workflow.
(5) AI-Generated The final product is largely created with AI; focus on prompting, editing, and evaluating AI output.

Why use this scale?
It enables transparency and flexibility. Not all courses or assignments need AI-integration, but many can benefit from clearly articulated expectations and alignment with learning outcomes. 

Mapping AI Use to Course Learning Outcomes

Start with Purpose
Before deciding where you assignment falls on the scale, ask:

  • What do I want students to learn from this task?
  • What skills, knowledge, or habits of mind does this demonstrate?
  • How might AI enhance–or undermine–that learning?

Aligning AI Levels with Fink's Categories of Significant Learning:

Fink Category Compatible AI Level Example

Foundational Knowledge

 2 to 3 (AI-Aware to AI-Optional) Students use AI to quiz themselves on concepts or generate questions.
Application 4 (AI-Integrated) Students compare their work to AI outputs to refine their strategies.
Integration 3 to 4 (AI-Optional to AI-Integrated) Students synthesize AI-generated ideas with other course materials.
Human Dimension 1 to 2 (AI-Free or AI-Aware) Students reflect on personal learning or ethical implications of AI. 
Caring Any level Assignments invite personal investment, regardless of AI use.
Learning How To Learn 5 (AI-Generated) Students critically evaluate AI tools and refine prompting techniques. 

Assignment Examples by Level

  1. AI-Free (e.g. Reflective Essay):
    Write a personal narrative about a time you experienced failure and what you learned. Do not use AI tools for any part of this assignment.
     
  2. AI-Aware (e.g. Research Summary):
    You may use AI tools to brainstorm or summarize articles, but you must cite any assistance and reflect briefly on its accuracy.
     
  3. AI-Optional (e.g. Design Proposal):
    You can choose to generate initial design drafts using AI. If you do, submit both the prompt and original output alongside your final version.
     
  4. AI-Integrated (e.g. Creative Writing Workshop):
    Use AI to generate a story based on a theme. Then rewrite and expand it, noting what changes you made and why.
     
  5. AI-Generated (e.g. Prompt Engineering Assignment):
    Your task is to create an AI-generated marketing pitch using a series of increasingly refined prompts. Your grade is based on process, critique, and strategy.