Skip to main content

Rethinking Student Evaluation in the Age of AI: A Call for Creative Pedagogy

The rapid rise of generative AI tools—such as ChatGPT, Gemini, Copilot, and others—is no longer a matter of future speculation; it is a present-day reality. These tools are fast becoming as common as calculators, smartphones, and personal computers once were. Just like we no longer question the presence of Microsoft Word in essay writing or the use of MATLAB in engineering computation, it is time we stop resisting the presence of AI in student learning. The question is no longer if AI should be accepted in education, but how we, as educators, can adapt to it in meaningful and constructive ways.

Beyond Detection: The Futility of Policing AI Usage

A segment of educators, understandably concerned with academic integrity, have begun relying on AI detectors to identify AI-generated content. While well-intentioned, this approach is reactive and largely unsustainable. AI detectors are imperfect, prone to false positives, and often become outdated as new models emerge. Worse still, this strategy may inadvertently discourage students from exploring powerful new tools that, if guided correctly, could deepen their understanding and enhance their learning journey.

Rather than policing AI, perhaps it is time to embrace it—and more importantly, to design evaluations that are AI-resilient.

Understanding the Complexity of Learning and Evaluation

Education has never been a one-size-fits-all process. We have always distinguished between:

  • Observed learning: Direct classroom instruction.

  • Self-directed learning: Independent study and exploration beyond the classroom.

  • Conditional evaluation: Exams under strict conditions.

  • Output-based evaluation: Grading based solely on submitted work.

  • Process-based evaluation: Assessments that consider progress, reflection, and methodology.

Each of these forms has its own merit. But in an AI-rich environment, output-based evaluation is the most vulnerable. When students are assessed solely based on their final submission, without insight into their process, challenges, or learning moments, it becomes increasingly difficult to distinguish between genuine understanding and AI-generated mimicry.

In contrast, project-based learning, process logs, in-class engagement, and real-world applications (e.g., interviews, live presentations, public dissemination) can provide richer, more authentic indicators of student growth.

The Misleading Comparison to “Back to Basics”

Some might argue that the fundamentals—reading, writing, arithmetic—must be reinforced, and rightly so. However, let us not conflate this with a rejection of technology. A child learning to count should not use a calculator. But a university student solving a multivariable differential equation should use computational tools. Higher education is not just about remembering content; it’s about applying, analyzing, evaluating, and creating.

The same logic applies to AI. Blocking its use may make sense at the early stages of skill formation, but restricting it in higher education only stifles innovation and curiosity. Just as medical students must know the theory of pathology, it is their bedside manner, real-world judgment, and surgical skill that define them in practice. The same is true for engineering, business, arts, and most certainly for Technical and Vocational Education and Training (TVET).

AI as a Companion, Not a Shortcut

Let’s reframe AI not as a crutch, but as a collaborator—an intelligent assistant that can support ideation, experimentation, and exploration. It is no different from how Grammarly helps students with their writing or how YouTube tutorials help learners grasp complex software.

But this only works if we redesign tasks:

  • Assign interviews that require emotional intelligence and real-world interaction.

  • Let students create public blogs, vlogs, or infographics where feedback comes from beyond the lecturer.

  • Incorporate community engagement, entrepreneurship, or prototype testing.

  • Use versioning tools like Git to track progress over time.

When students are encouraged to show how they think, not just what they produce, AI becomes a tool in their toolkit, not a substitute for effort.

A Final Word: Accept Early, Adapt Better

The integration of AI into education is inevitable. Those who accept and adapt early will shape the future of learning; those who resist may find themselves left behind. AI does not invalidate the role of educators—it redefines it. Our challenge now is not to detect AI-generated work but to reimagine the learning ecosystem in which students use AI ethically, creatively, and intelligently.

The Ministry of Higher Education Malaysia has issued the Guidelines on the Use of Generative Artificial Intelligence Technology in Teaching and Learning in Higher Education, indicating the readiness of educational institutions to integrate AI into the teaching and learning process. Link

As educators, our greatest responsibility is to prepare students for the real world, not just our classrooms. And in the real world, AI is already here.