Introduction

The panel “When to Embrace, When to Resist, and What’s at Stake with Generative AI” brought together faculty members from across the university to share perspectives on how generative AI is reshaping higher education. The event took place online via Zoom on 15 August 2025. Moderated by Soo Yuen Jien (Centre for Teaching, Learning and Technology [CTLT] and the School of Computing), the panel featured Alex Mitchell (Department of Communications and New Media), Vik Gopal (Department of Statistics and Data Science), and Shobha Avadhani (Department of Communications and New Media). The discussion combined faculty reflections, small-group breakouts, and audience exchange, highlighting both the promise and the tensions surrounding AI in teaching and learning.

Panel Perspectives

Building Foundations Before Turning to AI

Alex emphasised the need for students to develop essential foundational skills before using AI tools. While acknowledging that AI is unavoidable, he argued that students must still “go through the hard stuff.” Too often, he observed, students turn to AI because of workload pressures. Instead, course design should encourage intrinsic motivation and the enjoyment of learning, allowing students space to think critically and creatively rather than defaulting to efficiency alone.

Productivity, Employability, and Authentic Learning

Vik stressed discernment in how and when AI is used. While he acknowledged that GenAI can be a valuable aid in teaching preparation and productivity, he cautioned against using it for convenience or novelty. Overuse, he noted, risks depriving students of the “light-bulb moments” that come from grappling with complex problems. To safeguard authentic learning, he suggested setting explicit rules around AI use in group work and incorporating in-class, effort-based assessments. Looking beyond the classroom, he highlighted employability concerns: as technology advances rapidly, higher education risks being sidelined if it does not play a stronger role in shaping how graduates adapt to AI-driven workplaces.

Equity, Ethics, and the Relational Core of Education

Shobha framed the discussion within broader ethical and societal contexts. She argued that education is fundamentally relational, built on dialogue, mentorship, and human connection—elements that AI cannot replace. She urged faculty to interrogate the rhetoric of inevitability that often accompanies AI adoption, noting that corporate interests frequently drive this narrative. From a justice-oriented perspective, she advocated for empowering students with agency, enabling them to choose the tools they use while recognising that convenience should not override conscience.

Safeguarding Disciplinary Competencies

Drawing on his computing background, moderator Yuen Jien emphasised that AI literacy is important, but cannot substitute for core disciplinary skills. Computing students, for instance, must still learn to debug, verify, and evaluate code rather than relying solely on AI-generated outputs. He suggested that assessments in such fields should remain hands-on, proctored, and authentic to ensure students develop robust core skills and abilities.

Themes from Breakout Groups

Efficiency is Not Learning

Participants cautioned against equating efficiency with genuine understanding. Fluency in using AI outputs does not guarantee deep learning. For some, GenAI served as a “wake-up call,” reminding educators that students must still engage in higher-order thinking such as evaluative judgement, rather than focusing only on producing correct answers.

Employability and Future Readiness

Debates also centred on how AI is reshaping expectations for graduates. Some observed that the workplace now demands more complex skills as AI handles routine tasks. Others expressed concern that heavy reliance on AI could erode originality and reduce students’ capacity to generate new ideas. These conversations led to calls for revisiting learning outcomes to ensure higher education supports both career readiness and broader intellectual development.

Equity and Access

Equity emerged as a recurring concern. Unequal access to AI tools—whether paid or free—was viewed as a fairness issue. Groups also suggested that rubrics and assessments may need to be rethought in light of AI’s influence. Foundational learning was reinforced as a principle: novices should limit AI use until skills are secure, while more advanced learners might use AI to offload routine tasks.

The Relational Core of Education

Participants emphasised that education must remain relational. Although AI can simulate dialogue, students must understand that they are not interacting with a person. Safeguards are needed to help students develop empathy, communication skills, and their own voice in an AI-mediated environment.

Closing Reflections

The panel and group discussions presented a spectrum of views: from delaying AI use until students have solid foundations, to adopting it within clear boundaries, to questioning the structural assumptions driving its adoption. Despite these differences, there was broad agreement that GenAI should be integrated intentionally—embraced where it enhances learning, resisted where it undermines human, relational, or disciplinary values. Participants also noted that discipline-specific approaches will be essential as each field determines how best to engage with GenAI.

As GenAI becomes woven into higher education, how will you discern whether its use deepens your students’ learning—or whether holding back ultimately protects the very skills and capacities they need most?