Assessment, feedback and AI in online and distance education

Contrary to popular belief, contemporary discussion of assessment in higher education is not exclusively the domain of artificial intelligence (AI) evangelists, nay-sayers and everyone else in between trying to make sense of it. No, assessment as a fundamental pillar of teaching and learning has many facets, so it is refreshing to have a little bit of time dedicated to exploring the purpose of assessment. This is not just contextualised with the disruption of generative-AI, where students can generate a passable essay with a suitable amount of prompt engineering, but with the very principles of education, its relevance to individuals, employers and society at large.

This post summarises a few themes that caught my attention in breakout sessions by educators and a panel discussion led by Ishan Kolhatkar at the 2026 Research in Distance Education Conference, with expert contributions in higher education, assessment and online education. I’ve written before about the need for reframing formal education from the needs of the institution to the needs of the learner (Cornock, 2026). It’s with this perspective that I’ve noticed a similar shift in the discourse of assessment away from institutional process, competitive metrics and ensuring learners are meeting the prescribed outcomes of a course, towards assessment as a learning process, with alignment to post-course outcomes, professional learning and skills requires for lifelong learning.

Authentic assessment

Stewart Utley presented on authentic assessment, which provided a useful framing for how assessment is understood from a design perspective. In Utley’s context, the subject matter, a specific branch of medicine, changes regularly due to cutting-edge discipline research. Within that context there is clear alignment to professional practice. Indeed, the courses are pitched as professional development, rather than credit bearing programmes. Drawing on concepts relating to the subjectivity of authenticity (Ajjawi, et al., 2024; Gulikers, et al., 2004), Utley’s proposition is that authentic assessment is on a spectrum, and I think this is such a valuable contribution to move us beyond dichotomous descriptions.

“Theoretical: No real world connection.

Contextualised: Realistic scenario or case.

Representative: Mirrors professional tasks.

Genuine: Real audience and stakes.”

Utley (2026)

Moving beyond course outcomes

There is a significant difference between tasks that are replications of professional activities and genuine activities that come with authentic risks and consequences. In assessment, most risks are associated with the individual learner (will they pass the assessment), but authentic situations have a wider range of implications and in some settings can be financial, reputation or safety related. A truly authentic assessment therefore carries authentic risks. Utley’s premise is to “shift from replicating practices to building capabilities”, so that as students move into professional environments they are also able to adapt their practices. There is no faux authenticity.

Prof Emeritus Patricia Broadfoot stated during the panel discussion, “assessment is a blunt instrument” (especially examinations), and hence I infer a poor measure of competence to adapt to unknown situations as it is a snapshot in time of an individual’s learning. Assessment does not have to just be for the purposes of grades at the conclusion of a course. Traditionally, assessment as the culmination of a student’s work and understanding on a course, often with substantial time allocated, is framed for external measurement. Yet learning through assessment, with learning purpose focused on the individual, is an integral part of competency development. In subject areas that rapidly change, it makes pedagogic sense to focus on the competences to respond to changing knowledge, rather than application of knowledge at a specific point in time. In the panel discussion, Prof Chie Adachi emphasised the need for assessment to align to outcomes, not just module or programme outcomes, but graduate outcomes and by extension professional outcomes. Utley’s framing leads thinking forward to outcomes yet to be defined, which begins to shift thinking to how assessment is relevant to, and supporting of, learners as individuals.

Artificial intelligence for relevant assessment

In response to whether AI will transform or undermine assessment, Broadfoot replied, “Both!”. Without delving into often repeated debates, Broadfoot offered a fresh consideration of how AI could help students with the use of AI to support assessment of themselves as individuals, to help them prepare for lifelong learning by identifying gaps, opportunities and pathways. The relevancy of assessment through focusing on the individual is particularly interesting to shape the purpose of assessment-like activities, leading to, as Utley suggested, the use of generative AI by students for completing work as more obviously viewed as students only cheating themselves. In a later presentation by Prof Inés Gil-Jaurena, one of the student responses seemed to acknowledge the potential negative impact of AI use for their own studies: “Generative AI limits the development of mental capabilities, such as creativity and implications of new knowledge.” Students may indeed be aware that learning requires effort(!).

The potential for mixed messages between institutional AI policy against use, institutional capabilities that provide AI and professional/industry practice was noted in a later session by Prof Philip Powell and colleagues. Earlier, Adachi suggested a way to navigate this dissonance, by educators opening up discussions with trust and transparency about how AI and generative AI works for students in their context. For example, exploring how AI is used in the discipline, and how this relates or is different from academic work.

I would argue this confusion of policy and the distinction of AI in professional versus academic contexts applies not just to students, but to staff as well. Perhaps it’s time for some transparent and trustful conversations about AI use within and across academic teams and professional services. There is great opportunity there to really critically appraise existing quality assurance, assessment and operational policies and practices, which will require a better understanding of AI use in context.

Process over product

One of the often suggested approaches to combating academic misconduct with generative AI, is the use of formative assessment and assessment of process, rather than (just) final output. With authentic assessment, there is concern that a fully professional or work-based aligned output could far exceed what is needed for the academic assessment. To tackle this, Utley offered the concept of “Minimum Viable Professional Artefact”, a distillation of an authentic professional output into the core components that would be needed to evidence learning outcomes. For example, a sub section of a formal structured report, or evidence of a process. This seems to be a valid compromise of authenticity towards what is just enough to meet assessment needs, and aligns very well with Broadfoot’s call to check the “fitness for purpose” of assessment. 

Prof Naomi Winstone elaborated further, emphasising assessment of process, not just product. Winstone challenged that academic assessment itself is never an authentic workplace task, and I infer the institutional educational wrapper, including rubrics, grades and structured feedback also falling outside the authentic parameters. Instead of authenticity of task (or output), educators can look towards authenticity of the process to create those outputs. Process over product resonated with delegates as it formed a side-discussion later in the day in how formative assessment which aligns with summative outputs (as authentic scaffolding and assessment of process) is incongruous with institutional policy that demands anonymous assessment. Mixed messages, expectations and conflicting educational values abound! 

“It makes little sense to value a polished piece of work.”

Winstone (2026)

The (real) purpose of feedback

I found Winstone’s discussion of feedback particularly useful in highlighting that it is the process of feedback as a human exchange that drives learning, not the volume of comments provided to students. Noting how AI has been touted as a way to grade coursework, Winstone’s point is that feedback requires trust where educators can provide and discuss “feedback with awareness of the backstory of students”. AI may well be useful to provide comments on a student’s work based on a rubric, but how those comments are interpreted relates to where that student is on their individual learning journey, and that requires nuanced interaction. In the subsequent session led by Powell and colleagues, the risks of “inconsistent grading”, “bias against culturally diverse” assessment responses and “linguistically varied student submissions” are evident in automated, AI marking. However, such issues are also possible in human marking, perhaps due to level of experience in assessing. There are guardrails in place for the human version, such as moderation, second marking, discussion and appeal. However I, like others, am not convinced the guardrails are yet understood or established in an AI space that (in the worst case) removes humans from the process entirely.

Related to the idea of assessment and feedback as a dialogical partnership, near the end of the panel discussion there was a delightful dissection of the NSS (the UK National Student Survey of final year undergraduate students). The question set reinforces a somewhat outdated idea of students as recipients of learning, rather than participants of learning. For learning, feedback has to go beyond just being received by learners. As Winstone articulated, there requires engagement with feedback and opportunities to support and scaffold that engagement. 

Change is coming

In the final thoughts from the panel, Broadfoot set the tone for how I would like to view assessment and feedback within teaching and learning in the future:

“Change is coming. Revolutionary change is coming to society. We need to think about how education will change… Putting students in the driving seat with self-assessment and feedback… We need students to say, ‘I’m weak on this, I need to learn that.’”

Broadfoot (2026)

To me, this describes an educational system that is, as Broadfoot suggested earlier, not of competition, but of competency. Competency requires a maturity of understanding about your own strengths and weaknesses, where and how you can improve, self-assessment and self-reflection. Education (from the very start of formal education) is not about ‘knowing stuff’, but there is, arguably, a societal norm that values knowledge as an output of learning, rather than application and knowing. 

As higher education enters a phase of change, this is an opportunity to really shape an educational system that remains grounded in academic rigour, critical thinking and creativity, but also adapts, enabling greater responsiveness and inclusivity. Assessment for learning is not a new idea. This reframing of assessment for ongoing learning, with action, can enable all learners to become equipped to have greater ownership over their own education by understanding their own learner needs better. For those of us who are self-critical, we should channel that self-assessment recognising we are each on a lifelong learning journey.

References 

1. Ajjawi, R., Tai, J., Dollinger, M., Dawson, P., Boud, D. and Bearman, M. (2024). ‘From authentic assessment to authenticity in assessment: broadening perspectives’, Assessment and Evaluation in Higher Education, 49(4), 499-510.
2. Cornock, M. (2026). Institutional readiness for the future of education. Hatzipanagos, S., Brown, S., Powell, P., Milner, M., Zain, N., Hewawasam, M. (2026). Generative AI and the transformation of assessment: a mixed methods study of practice, perceptions and policy. 20th Annual International Research in Distance Education Conference (RIDE2026). London. 12-13 March 2026.
3. Gil-Jaurena, I. (2026). Supporting students’ learning through renewable assignments in distance education. 20th Annual International Research in Distance Education Conference (RIDE2026). London. 12-13 March 2026.
4. Gulikers, J. T. M., Bastiaens, T. J. and Kirschner, P. A. (2004). ‘A five-dimensional framework for authentic assessment’, Educational Technology Research and Development, 52(3), 67-86.
5. Kolhatkar, I., Adachi, C., Broadfoot, P. and Winstone, N. (2026). Envisioning the future of assessment. Panel. 20th Annual International Research in Distance Education Conference (RIDE2026). London. 12-13 March 2026.
6. Utley, S. (2026). Authentic assessment in short- form online courses: A case study. 20th Annual International Research in Distance Education Conference (RIDE2026). London. 12-13 March 2026.