Post panic and prohibition: An Asian university’s AI responses

Universities across the world are grappling with the implications of generative artificial intelligence (GenAI). Tools such as ChatGPT, Claude and NotebookLM have unsettled long-standing assumptions about teaching, assessment, writing and intellectual development. The responses within universities have therefore been deeply emotional and uncertain.

Fear, anxiety, scepticism, excitement and fascination coexist uneasily within academic spaces, with many academics simultaneously using AI in their own work while worrying about its effects on student learning and intellectual formation.

As part of my research and benchmarking work on AI and higher education across South Africa, Africa and broader Global South contexts, I recently visited three of Southeast Asia’s leading world-ranked universities to explore how they are responding institutionally and pedagogically to the AI challenge.

This article draws particularly on learnings from one of these universities, whose responses illuminate the productive but deeply complex challenges universities are now navigating.

This uncertainty is unfolding at a moment when universities globally are already under pressure from political polarisation, economic instrumentalism and intensifying attacks on academic freedom and the public purposes of higher education.

AI is therefore not arriving in stable institutional environments. Rather, it is amplifying existing tensions around knowledge, authority, democracy and the role of universities in society.

AI as disruption and uncertainty

During April 2026, I conducted a series of conversation-style engagements with academic leaders, lecturers, teaching and learning specialists, curriculum planners and strategic planning staff at one of these universities.

These discussions were deliberately dialogical rather than structured interviews. The intention was not to extract fixed institutional positions but to understand how academics themselves were making sense of the rapidly shifting AI landscape.

What emerged at this university was not institutional certainty but a mixture of anxiety, experimentation, curiosity, caution and pedagogical reflection. Academics repeatedly described feeling simultaneously excited by AI’s possibilities and unsettled by its implications for learning, assessment, authorship and intellectual formation.

One of the most striking aspects of the university’s response was that institutional leaders recognised relatively early that GenAI represented more than simply another digital tool to be integrated into existing systems. Academics repeatedly described AI as disrupting foundational assumptions underpinning modern university education.

Writing, long treated as evidence of reasoning and intellectual development, can now increasingly be generated by machines.

Assessment systems built around text production no longer operate with the same certainty, and lecturers cannot always easily determine where student thinking ends and AI-generated output begins.

This has produced both practical and emotional turbulence within the institution. Some academics expressed concern regarding the erosion of intellectual scaffolding processes through which students historically developed understanding. Others worry that students increasingly bypass difficult cognitive work by using instant AI-generated answers.

Yet alongside these anxieties was also curiosity and experimentation.

Several academics described AI as forcing universities to confront educational questions that perhaps should have been addressed long ago: What actually constitutes meaningful learning? What should assessments evaluate? Is writing itself equivalent to understanding? What forms of knowledge and judgement matter most in contemporary society?

These questions have no settled answers, which is precisely why the current moment feels so unsettled.

Governing amid uncertainty

Importantly, the university did not respond to AI through blanket prohibition or technological surrender. Instead, it attempted to create a negotiated institutional environment within which different perspectives and responses could coexist.

The institution developed university-wide AI guidelines and broader policy frameworks through consultative processes involving academic staff, teaching and learning specialists, planners and governance structures.

What emerged, however, was not a rigid compliance regime. Rather, the university appears to have created enabling conditions for ongoing experimentation, adaptation and debate. Regulation functions less as a mechanism of closure than as a framework for collectively navigating uncertainty.

This may be one of the most important lessons emerging from the institution’s experience. Much popular discussion frames regulation and innovation as oppositional. Yet the university demonstrates that carefully constructed governance environments can create space for pedagogical experimentation while still providing institutional coherence.

At the same time, institutional leaders openly acknowledged that many responses remain provisional and unresolved. Assessment systems remain unstable. AI detection tools are increasingly unreliable. Academic staff differ significantly in their orientations and capabilities.

Faculty members continue to negotiate their own disciplinary responses, while policy frameworks themselves are evolving as technology changes. The university therefore embodies a condition increasingly visible across global higher education: institutions are governing amid uncertainty rather than beyond it.

Different disciplines, different anxieties

Another important insight emerging from the visit was that AI does not enter all disciplines in the same way. Different faculties are navigating different forms of tension, opportunity and vulnerability shaped by their own disciplinary cultures and professional traditions.

In education-related fields, discussions were often deeply philosophical and pedagogical. Academics reflected on how AI destabilises assumptions about writing, thinking and intellectual development itself. One senior academic remarked that universities may be entering an “era of post-authorship”, in which written work can no longer reliably serve as evidence of student understanding.

This has generated both anxiety and innovation. Some lecturers increasingly use oral presentations, collaborative classroom activities, reflective exercises and project-based learning to assess conceptual understanding. Students may be asked to explain how they used AI, justify their prompts or verbally defend their written submissions.

These approaches attempt to move beyond simply evaluating outputs toward examining intellectual engagement and reasoning processes.

Yet academics also acknowledged the practical difficulties involved. Large classes, limited staff capacity and heavy workloads complicate more labour-intensive forms of assessment. There is therefore no straightforward pedagogical solution, only ongoing experimentation and adjustment.

In healthcare-related fields, AI has raised different concerns. Academics described how AI can support diagnostic learning and patient communication training, while simultaneously expressing concern that overreliance on AI-mediated interaction could erode empathy, trust and the human dimensions of professional practice.

Engineering and technical disciplines tended to frame AI somewhat differently. Here, AI was viewed as a potentially useful tool for report writing, visualisation and data analysis while still requiring strong disciplinary judgement and conceptual interpretation.

Meanwhile, academics in humanities-related fields expressed concern that AI may accelerate broader global trends marginalising the humanities in favour of narrowly instrumental and economically driven educational priorities. What became evident across these discussions is that AI is not simply a technological issue. It is also epistemological, professional, ethical and deeply human.

The reorganisation of intellectual labour

The challenge confronting universities is not merely the presence of tools such as ChatGPT, Claude and NotebookLM but the wider reorganisation of intellectual labour, reading, writing and knowledge production that these platforms are accelerating.

Students now move fluidly between multiple AI systems for summarisation, drafting, conceptual explanation, note synthesis and research support. AI increasingly accompanies students during lectures, tutorials, assignment preparation and study sessions.

Academics acknowledged that these technologies are no longer peripheral to student learning practices. They are becoming embedded within the everyday rhythms of university learning itself. This is creating uncertainty about which intellectual work students should still perform themselves and which forms of AI support are educationally legitimate.

The concern is not simply that students may ‘cheat’. The deeper issue is whether AI may gradually weaken the slow intellectual processes through which understanding historically developed: reading closely, drafting imperfectly, revising arguments, struggling conceptually and learning through cognitive difficulty.

At the same time, many academics recognised that AI can support learning meaningfully when used critically and reflectively.

The challenge, therefore, lies not in eliminating AI from education, which is increasingly impossible, but in redesigning educational processes that can preserve intellectual formation within AI-mediated environments.

The assessment crisis

Assessment emerged repeatedly as the university’s central institutional challenge. Like universities globally, the institution is grappling with how to evaluate learning in contexts where AI can rapidly generate sophisticated text and responses.

Importantly, many academics no longer believe that surveillance and detection alone can solve the problem. Several openly acknowledged the limitations of AI-detection software and the futility of relying exclusively on policing approaches. Instead, the university is experimenting with more transparent and negotiated forms of governance.

Students are increasingly required to declare whether and how they used AI in assignments. Lecturers similarly disclose their own use of AI in teaching preparation. Assessment categories range from no permitted use of AI to fully AI-integrated tasks. This dual transparency model shifts the conversation away from concealment and punishment toward ethical engagement and accountability.

Yet tensions remain unresolved. Some academics remain resistant to students’ use of AI, while others view responsible AI engagement as essential preparation for professional life. Many occupy uncertain middle positions. The institution therefore continues to navigate contradictory pressures: preserving academic integrity while recognising that AI is now deeply embedded in everyday learning and professional environments.

Knowledge reproduction vs knowledge-making

One of the most conceptually significant themes emerging from the university’s response was a growing critique of educational models centred primarily on knowledge reproduction. Several academics argued that AI now performs many reproductive educational functions remarkably well. It can summarise information, synthesise texts and generate structured responses almost instantaneously.

This creates discomfort for institutions historically organised around information transmission and written reproduction. Yet rather than concluding that universities have become irrelevant, many academics argued that AI may force universities to rediscover deeper educational purposes.

Increasingly, discussions revolve around cultivating judgement, conceptual understanding, ethical reasoning, collaboration and reflective engagement rather than merely reproducing information. There are early signs of movement toward more design-based and process-orientated pedagogies, although these shifts remain partial and contested.

This is particularly relevant within African higher education contexts.

My own research across South African universities suggests that institutions are grappling with similar tensions. Universities are simultaneously trying to prepare students for AI-mediated futures while preserving critical intellectual traditions, disciplinary depth and human agency. The challenge is especially acute in contexts marked by inequality, uneven digital access and resource constraints.

The human question

Perhaps the deepest issue confronting this university concerns the future purpose of higher education itself. The academics I spoke to worry that AI risks intensifying broader global trends in which universities become increasingly orientated toward efficiency, optimisation and labour-market responsiveness at the expense of civic, ethical and intellectual formation.

One academic reflected that universities may need to shift from functioning primarily as information gatekeepers toward becoming developmental institutions focused on cultivating thinking, judgement and disciplined understanding. This observation carries major implications for universities globally, particularly within the Global South.

Across Africa and elsewhere, higher education institutions face mounting pressure to produce technologically adaptable graduates for rapidly changing economies. Yet universities also remain crucial public institutions responsible for cultivating democratic participation, ethical reasoning and social responsibility.

The AI challenge, therefore, cannot be reduced to technological adaptation alone. It is fundamentally about how universities preserve meaningful forms of human formation within increasingly platformised and AI-mediated societies.

One participant posed a deceptively simple question that lingered throughout my visit: “How does what we are doing relate to being human?” That question sits at the centre of the contemporary AI debate.

A productive but unfinished experiment

The university I visited has not solved the AI problem, nor would it claim to have done so. Its response remains unfinished, experimental and internally contested. Yet this may precisely be what makes it instructive.

The university demonstrates that institutions need not choose between technological surrender and reactionary resistance. It is possible to build adaptive institutional ecosystems that combine regulation with experimentation, ethical reflection with innovation and institutional coordination with disciplinary flexibility.

Most importantly, the institution recognises that there are no final or stable answers available yet. The AI challenge is unfolding faster than universities can fully comprehend. Higher education institutions everywhere are therefore navigating uncertainty, affective turbulence and profound educational change simultaneously.

For African universities now grappling with similar realities, this lesson may be especially important. The challenge is not merely how to incorporate AI into existing university systems. It is about how to rethink teaching, learning, assessment, knowledge and human formation in profoundly changing times while holding on to the human purposes of higher education itself.

Aslam Fataar is a research and development professor in higher education transformation in the department of education policy studies at Stellenbosch University in South Africa.

This article is a commentary. Commentary articles are the opinions of the author and do not necessarily reflect the views of University World News.