Two years ago, I published an article introducing a framework for AI and university assessment, aiming to clarify opportunities, challenges, and implementation strategies. This framework has repeatedly proven valuable in discussions with universities seeking to modernise their assessment practices. Yet, despite rapid technological advances, particularly in generative AI, and evolving workplace demands, the higher education sector has struggled to act swiftly and decisively. Recent sector events I have attended have continually echoed familiar calls for change but rarely offer tangible progress or practical solutions to rebuild assessment. Worryingly, some institutions are reverting to outdated approaches like proctored exams and live vivas, suggesting a retrofit than the new-build that is now needed.
Beyond Exams: Why They Are Not the Answer
Online proctored exams, which surged during the COVID-19 pandemic, have highlighted cracks in reliability and validity, while exam-based models—digital or not—generally overemphasise memorisation instead of critical thinking, creativity, and adaptability. Time-pressured exams in online or hybrid contexts can worsen academic misconduct, heighten student anxiety, and disadvantage those lacking reliable technology (Woldeab & Brothen, 2019; Hartnett et al., 2023). In the case of fully online learners from around the globe juggling multiple commitments, rigid testing schedules pose yet another barrier. Meanwhile, AI tools are becoming ever more accessible, making the quest for fair, inclusive, meaningful assessment more urgent than ever — spurring the push for alternatives to traditional exams.
From Continuous to Authentic Assessment: Meaningful Assessment Pathways
Continuous assessment, which breaks learning into a series of smaller tasks spread out across a module or program, has been advocated by educationalists as an effective alternative for the past few decades. Instead of one final exam or essay submsision, students engage in frequent formative activities—quizzes, discussion posts, or short reflections—alongside summative assessments such as projects or presentations. This approach allows instructors to catch misconceptions early and provide regular feedback. Students likewise experience lower pressure, better work–life balance, and more opportunities to prove their mastery over time (Fynn & Mashile, 2022). The key is designing enough checkpoints for meaningful feedback without overburdening staff or learners. Thinking more broadly about feedback mechanisms is key here – not everything has to be tutor-led. Cohort feedback, peer feedback, automated feedback, self-feedback can all play a part, being clear in student instructions about what form feedback will take being essential to set expectations.
Authentic assessment, is also an important part of our learning design toolkit. This approach requires students to tackle tasks that mirror real-world demands. Rather than testing recall of course content, authentic assessments ask learners to synthesise and apply their knowledge in practical ways—developing marketing campaigns, conducting real-time data analysis, or drafting policy briefs. These methods require higher-order thinking, collaborative problem-solving, and creativity, more accurately reflecting the workplace challenges graduates will face (Vlachopoulos & Makri, 2024). Rather than creating more complex assignments, simply shifting writing genres and the types of assessment artefacts we require students to produce is an easy switch. By bridging theory and practice, authentic assessment also reduces incentives for cheating, because the task itself is both engaging and personalised.
The AI Factor: Challenges and Opportunities
With generative AI tools (e.g., ChatGPT, CoPilot and Claude) now at students’ fingertips, educators must rethink assessment design to incorporate this technology constructively. Embedding AI use into assessment needs to move beyond merely testing whether students can spot ‘hallucinations’ or factual errors in AI outputs. As generative AI technology evolves, such exercises risk becoming outdated, focusing on superficial AI literacy rather than deep disciplinary thinking. Recent research from the Higher Education Policy Institute (HEPI) and Kortex reported that in 2025 88% of students surveyed used AI tools to develop their assessments, and increase of 35% from the previous year (Freeman, J. 2025). Merely banning AI is rarely practical; instead, we should clarify ethical guidelines around AI usage, for instance by using the the Furze AI Assessment Scale (Perkins et al., 2024; Furze, 2024). Students can be asked to show how AI informed their thinking or assisted data analysis, while still demonstrating original, critical judgment. If they do use AI to streamline tasks, they must cite it properly and explain where human insight added value.
In speaking with sector colleagues, it’s clear that AI detection tools often add workload and foster mistrust, shifting focus from evaluating learning to policing students—especially when detections are unreliable. I firmly believe we should trust our students and empower them to make informed choices about AI use. Our role is to design assessments that minimise misconduct and promote integrity through engaging, reflective learning. If programme leads are concerned that an assessment is at high risk of being completed by AI they can use a tool such as the AI Risk Management Scale (ARMS) to support the redesign of the tasks.
Building the Bigger Picture: Programme Level Assessment
As AI capabilities advance and remote learning becomes more common, institutions must explore alternative approaches that better capture the critical thinking, creativity, and adaptability graduates need in today’s fast-changing world. When university education became more modular, we created a structure that instantly increased assessment burden and started to erode the programme-experience.
Programme-level assessment connects multiple modules into a cohesive, project-driven framework. Instead of isolating learning outcomes within individual courses, it emphasises students’ cumulative growth, culminating in capstones or e-portfolios that demonstrate authentic skills. This approach not only reveals deeper integration of knowledge but also reduces assessment workload for students and faculty alike. The emergence of flexible VLE plugins helps simplify the shift away from rigid module structures, fostering a more holistic programme engagement. By embracing pedagogies from creative and reflective disciplines, as well as apprenticsehip approaches, programme-level assessment crucially supports human agency within modern assessment practices. As such I this is a shift we should seriously consider.
Conclusion
In an era where technology and professional expectations evolve at breakneck speed, building continuous, authentic, and AI-ready assessment strategies is no longer optional—it’s essential. By trading in traditional, surveillance-heavy exams for tasks that engage real-world thinking, we promote genuine mastery and integrity. This reimagined approach ultimately prepares students for success beyond graduation, equipping them with the critical faculties, ethical mindsets, and adaptability they need to thrive in the modern world.
References
Freeman, J. (2025) Student Generative AI Survey 2025. HEPI Policy Note 61 (February 2025). Available at: https://www.hepi.ac.uk/2025/02/26/hepi-kortext-ai-survey-shows-explosive-increase-in-the-use-of-generative-ai-tools-by-students/ (Accessed: 23 March 2025).
Furze, L. (2024) ‘Updating the AI Assessment Scale’ [Blog] Leon Furze https://leonfurze.com/. Available at: https://leonfurze.com/2024/08/28/updating-the-ai-assessment-scale/ (Accessed: 23 March 2025).
Fynn, T. & Mashile, J. (2022) ‘Continuous online assessment at a South African ODeL institution’, Frontiersin Education, 7. Available at: https://doi.org/10.3389/feduc.2022.791271 (Accessed: 23 March 2025).
Hartnett, M., Butler, P. and Rawlins, P. (2023) ‘Online proctored exams and digital inequalities during the pandemic’, Journal of Computer Assisted Learning, 1 – 13. Available at: https://doi.org/10.1111/jcal.12813 (Accessed: 23 March 2025).
Perkins, M., Furze, L., Roe, J., & MacVaugh, J. (2024) The Artificial Intelligence Assessment Scale (AIAS): A Framework for Ethical Integration of Generative AI in Educational Assessment. Journal of University Teaching and Learning Practice, 21(06). Available at: https://doi.org/10.53761/q3azde36 (Accessed: 23 March 2025).
Vlachopoulos, P. & Makri, A. (2024) ‘A systematic literature review on authentic assessment in higher education: Best practices for the development of 21st century skills, and policy considerations’, Studies in Educational Evaluation, 83, Article 39. Available at: https://doi.org/10.1016/j.stueduc.2024.101425 (Accessed: 23 March 2025).
Woldeab, D. and Brothen, T. (2019) ‘21st Century Assessment: Online proctoring, test anxiety, and student performance’, International Journal or E-Learning & Distance Education, 34(1).. Available at: https://www.ijede.ca/index.php/jde/article/view/1106 (Accessed: 23 March 2025).
Kate Lindsay Blogs 