Artificial Intelligence (AI), particularly generative tools like ChatGPT, is no longer a futuristic abstraction – it is a potentially disruptive technology reshaping how we live, work and learn.
In higher education, assessments are among the most impacted domains. In a recent article, Integrating Artificial Intelligence into Higher Education Assessment, I propose a radical yet practical shift in the way universities can design, deliver and evaluate student assessments.
https://doi.org/10.61669/001c.131915
Rather than resisting AI, a re-calibration of assessment strategies may be needed, that not only embrace AI’s capabilities but use them to enhance authenticity, inclusivity and relevance in student learning.
Why Traditional Assessments Are Losing Ground
The longstanding “standard assessment paradigm” of essays, multiple-choice questions (MCQs) and short-answer questions (SAQs) are becoming increasingly susceptible to AI-generated content. These formats, while familiar and scalable, are quickly becoming obsolete for several reasons:
-
Inauthenticity: Traditional tasks often fail to simulate real-world challenges or professional contexts.
-
Equity concerns: Standardised formats can disadvantage diverse learners and limit how students can demonstrate their learning.
-
Vulnerability to AI: Generative models like ChatGPT can create high-scoring essays and answer exam questions, raising issues of academic integrity and validity.
There is therefore an urgent need to re-design assessment modalities for a digital and AI-driven world.
Case Study: Integrating AI into Postgraduate Coursework Assessment
I present a practical case study from a postgraduate immunology module, where students were asked to use ChatGPT to draft an outline for a standard coursework essay. This was supported by formative workshops in the responsible use of AI, followed by peer evaluation and final submission without further AI assistance.
Key Observations from the Study:
-
Student engagement: While 58% had used generative AI before, only 5% were confident in using it – highlighting a significant gap in AI literacy.
-
Utility and limitations: Students valued AI for brainstorming, structuring, and summarising. However, they criticised its lack of depth, accuracy issues and superficial output.
-
Views on plagiarism: Only 26% of students viewed using AI in assessments as plagiarism, revealing inconsistent understandings of academic integrity.
These findings confirm that while AI can enrich learning experiences, student and staff training is critical to ensure ethical and effective use.
Integrating AI: A New Framework for Assessment
Incorporating AI into coursework assessment can be dealt with in two ways:
1. Mitigation Strategies
Design assessments that are less vulnerable to AI misuse, such as:
-
Oral presentations and vivas
-
Laboratory and clinical evaluations
-
In-person, invigilated exams
-
Experiential tasks like fieldwork or practicals
2. Integration Strategies
Incorporate AI into the assessment design itself. For example:
-
Using AI to draft or co-create content (e.g., essays, code, visualisations)
-
Critiquing or refining AI-generated output
-
Engaging in debates or simulations with AI as a participant
-
Creating tasks where AI serves as a research assistant or idea generator
These approaches not only maintain academic rigour but also mirror how professionals increasingly engage with AI in the workplace.
The third way
There is of course a third option – ignore or completely avoid AI. In my view, this approach is less progressive and prevents opportunities to engage students with AI.
Incorporating AI into assessments has the benefit of training students on its limitations, capability and the ethical use of generative AI in their studies.
AI and Feedback: A New Frontier for Learning Support
Another potential opportunity for AI involves integrating its use into feedback practices. For example, AI may be useful for the following:
-
Provide immediate and personalised feedback
-
Suggest improvements for writing and structure
-
Generate rubrics, questions and learning tasks
-
Help students self-assess and reflect
However, caution is warranted. AI feedback may lack contextual awareness, discipline-specific insight or alignment with educator expectations. Therefore, human oversight remains essential.
Implications for Higher Education Policy and Practice
The integration of AI into assessments is not a temporary fix but a strategic necessity. Institutions must:
-
Invest in AI literacy training for both students and staff
-
Develop clear AI usage and ethics guidelines
-
Re-frame learning outcomes to include digital competencies
-
Emphasise authentic, higher-order assessments over rote tasks
Importantly, AI should support, not replace, human educators. Personal interaction, empathy and mentorship are irreplaceable elements of education.
Towards a Future-Ready Assessment Model
By embedding AI into coursework in thoughtful, transparent and pedagogically sound ways, universities can:
-
Promote critical thinking and creativity
-
Enhance authenticity and inclusivity
-
Safeguard academic integrity
-
Equip graduates for an AI-driven world
The question is no longer whether to integrate AI into assessments, but how best to do so.
As higher education faces the realities of AI disruption, universities should embrace AI integration.
It is possible to innovate and embrace a progressive culture, while protecting academic standards. Universities have a responsibility to train their students in responsible and proper generative AI use and incorporating AI into assessments will be essential for the future of learning.
References
You can read the original article here: https://doi.org/10.61669/001c.131915
Williams, A. (2025). Integrating Artificial Intelligence Into Higher Education Assessment. Intersection: A Journal at the Intersection of Assessment and Learning, 6(1), 128–154. https://doi.org/10.61669/001c.131915