Rethinking Assessment for Generative AI: Ungrading

This post is part of a series on rethinking assessment for generative AI. Check out the earlier posts below:

While Generative AI has thrown both K-12 and tertiary education into disarray, it’s time to move the narrative away from “cheating” and to start focusing on what we can do about it. This post explores the concept of “ungrading”, an approach to shift the attention away from final scores and back to the learning.

What is ungrading?

Ungrading is an approach that deviates from traditional grading systems, favouring a more feedback-centric model. Instead of focusing on scores or letter grades, the emphasis shifts towards providing detailed, constructive feedback, encouraging students to reflect on their learning and grow from their experiences.

Self-assessment and peer review are encouraged in ungrading methods. Alongside other teacher-led forms of feedback, students gain a better understanding of their own learning, the assessment criteria, and benefit from the diverse perspectives of their peers through reflection and critique.

Ungrading, as the name suggests, steers away from traditional letter and number grades. This could definitely help alleviate the fear and competition often associated with grading, and encourage more of the kinds of collaboration and team work we’re always saying we want in the classroom. One huge issue with our current models of assessment is that they get in the way of genuine interactions like this: it’s fine to say “work together”, but if the end-game is a ranking process (like the senior secondary ATAR in Australia), then there’s a contradiction.

Clear and transparent assessment criteria remains pivotal in ungrading. Guidelines and expectations are instrumental for effective assessment and meaningful feedback, irrespective of whether there’s a letter or number attached.

If you want a thorough definition of ungrading, with an exploration of how it has been used and some directions for the future, check out Jesse Stommel’s blog post here.

The live webinar Practical Strategies for Image Generation in Education is now available as a recording. Click here to purchase the recorded webinar.

Ungrading and GAI

The de-emphasis of competitive grades and a final number means that ungrading has a lot of potential to help with assessment and GAI. As I’ve written previously, there are many reasons students cheat: academic pressure and the need to perform are among them.

Emily Pitts Donahue, Associate Director of Instructional Support at University of Mississippi, and her students Abi and Trey explore the benefits and challenges of ungrading in the context of GAI on Pitts Donahue’s substack. Their story reveals the double-edged sword of GAI in assessment: something most of us are pretty familiar with by now. On one hand, GAI tools like ChatGPT can serve as learning aids, especially in assisting writing skills among students. On the other hand, these very tools can tempt students to circumvent the learning by outsourcing their thinking and writing tasks to AI.

Pitts Donahue labels an entrenched focus on grades as one of the core underlying issues driving students to engage in academic dishonesty. When the primary aim of education shifts towards attaining higher grades rather than gaining knowledge and honing skills, students are more likely to turn to GAI for completing their assignments. This grade-centric outlook not only undervalues the learning journey but also undermines the educational goals. It’s in this scenario that ungrading emerges as a potential antidote to the GAI-induced challenges in assessment.

Pitts Donahue’s classroom experiment with ungrading showed a promising shift in students’ approach towards learning. By eschewing points or letter grades and instead providing extensive written feedback with opportunities for revision, the ungrading approach redirected the focus towards learning and improvement.

This also led to meaningful dialogues with students about the appropriate use of GAI in the learning process. When instances of misuse of GAI arose, they became springboards for discussing how such misuse impeded learning rather than just affecting grades. The shift from a product-centred to a process-centred assessment reinstated the value in the learning process and helped students to engage authentically with the material, without the temptation for misusing of GAI.

tired black man lying on opened book and homework papers
Photo by Monstera Production on

Challenges of Ungrading

Transitioning from traditional grading systems to an ungrading approach obviously comes with a set of challenges. Here are some of the challenges associated with ungrading, drawn from various perspectives (references at end) and summarised with ChatGPT (Model: GPT-4 with Bing search):

  1. External Pressures and Self-Evaluation:
    • One of the challenges stems from external pressures, particularly when self-evaluation is a component of ungrading. Given that grades often play a crucial role in admissions to further education and job selections, the pressure on students to achieve high grades remains, even in an ungrading setup​.
    • Not all students are equally equipped to self-evaluate, and there’s a concern that some students might undervalue their work, while others might overvalue theirs. However, certain strategies like well-described rubrics can help mitigate potential biases in self-rating​.
  2. Misinterpretation and Misapplication:
    • The true essence of ungrading can easily be lost if not well-understood or well-implemented. A significant number of educators who attempt ungrading still rely on rubrics, stated learning outcomes, and other traditional grading elements, albeit under different terminologies. This misalignment with the core philosophy of ungrading doesn’t change the students’ assessment in any meaningful way, nor does it alleviate students’ fears or competitive pressures associated with grading​.
  3. Time-Consuming Feedback:
    • Providing detailed feedback, which is a cornerstone of ungrading, can be time-consuming, especially in large courses. The process demands a substantial amount of time and effort from educators to ensure meaningful feedback that can guide students towards better understanding and improvement​.
  4. Scaling Challenges:
    • Scaling the ungrading approach to larger classes or institutions with traditional grading ingrained in their systems poses a significant challenge. The logistical and cultural shift required for ungrading may face resistance or implementation hurdles.
  5. Lack of Clear Standards:
    • In ungrading, the absence of clear standards or specifications might cause ambiguity for both students and educators. Although some versions of ungrading like specifications grading attempt to address this by setting clear standards for each assignment, the broader practice of ungrading might face challenges in defining success and understanding progress​.
  6. Educational Culture and Mindset:
    • The entrenched culture of grading and the mindset associated with it can be significant barriers to the adoption of ungrading. Overcoming these cultural and psychological hurdles requires a concerted effort from educators, administrators, students, and the broader educational community.

Interested in more articles like this? Join the mailing list:

Success! You're on the list.

Where to start with ungrading

Transitioning to an ungrading system isn’t going to be easy. You’ll meet institutional pushback, resistance to change, and no doubt resistance from students who just want to know that final number or letter. The transition should be about creating a culture of trust, feedback, and continuous learning, which can then help in mitigating the potential misuse of GAI technologies.

  1. Transparency and Dialogue: Establish an open dialogue with students about the benefits and limitations of GAI through academic integrity conversations, and why/how/where GAI should and shouldn’t be used in the learning process.
  2. A Feedback-Rich Environment: Replacing grades with detailed feedback can help in shifting the focus from performance to actual understanding and improvement.
  3. Peer and Self-Assessment: Encouraging students to engage in peer reviews and self-assessments can promote a deeper understanding of the learning material and the assessment criteria.
  4. Specifications Grading: Consider specifications grading as a variation, which bundles assessments together. Clear and transparent assessment criteria should also be a part of ungrading, helping to maintain high educational standards while promoting authentic learning.
  5. Professional Development for Educators: Preparing educators for this shift is incredibly important. Training on how to provide effective feedback and how to engage students in self and peer assessments will be crucial for the whole faculty or organisation.

Ungrading can be just another tool in the suite of methods you use to rethink assessment for GAI. It’s been around since before GAI was on our radars, and there are plenty of studies out there exploring the benefits and challenges in both K-12 and tertiary contexts.

Alongside orals and discussions, alternative assessment forms to essays, and an understanding that Generative AI doesn’t automatically constitute cheating, ungrading could be a useful idea to carry into 2024.

If you’d like to get in touch to discuss GAI, academic integrity, or assessment, please contact me using the form below:


Nilson, L. B. (2016, January 18). New ways to grade more effectively (essay). Inside Higher Ed. Retrieved from

Stommel, J. (2023, April 6). What is Ungrading?: a Q&A. Jesse Stommel. Retrieved from

Morris, S. M. (2022, October 21). The problem with ungrading? Everyone’s doing it wrong. Times Higher Education. Retrieved from

Kenyon, A. (2022, September 21). What is Ungrading? Duke Learning Innovation. Retrieved from

Pitts Donahoe, E. (2023, September 29). The Rise of Generative AI Calls for New Approaches to Grading. Emily Pitts Donahoe Substack. Retrieved from

Leave a Reply

%d bloggers like this: