AI for Education: The Digital Transformation of a Liberal Arts Institution — Implementation at Lingnan University

Synthesis: Qin offers a position paper and institutional case analysis of how Lingnan University — Hong Kong's longest-established higher education institution — is strategically transforming itself into a "Research-Intensive Liberal Arts Institution in the Digital Era" without relinquishing its humanistic foundations. Framed through sociotechnical systems (STS) theory, the analysis spans four interrelated dimensions: upgrading instructional practice through AI, prioritizing irreplaceable human competencies, renewing curricula, and retaining ethical and cultural value systems. The paper argues that the GenAI revolution constitutes an intellectual rather than a technocentric transformation, positioning digital fluency as a core liberal arts competency and proposing a human-in-the-loop educational model as a blueprint for harmonizing technological innovation with enduring moral and intellectual aims.

Key Findings

• Framing AI as an intellectual revolution, not mechanization of physical labor. Because GenAI automates cognitive work, Lingnan repositions the human scholar and student from "laborer" to "architect or conductor" of knowledge, treating AI as a partner in intellect rather than a replacement — an orientation that underpins its curriculum restructuring and emphasis on human-centric skills.
• An "adoption-first," equity-driven policy democratizes GenAI access. Since August 2023 Lingnan has provided campus-wide, free access to advanced GenAI tools (ChatGPT and DeepSeek) for all students and faculty, reasoning that prohibition is futile and that digital fluency should not depend on a student's ability to pay — catalyzing bottom-up innovation among faculty in non-STEM disciplines like History, Philosophy, and Visual Arts.
• The GenAI Assessment System (GAAS) embodies human-in-the-loop assessment. AI analyzes student submissions against instructor-supplied rubrics, but its feedback is always routed first to the instructor for review and validation before reaching the student, who must then reflect and respond — addressing timeliness, consistency (uniform rubric application), and personalization, and freeing instructors for mentorship-style interactions.
• AI literacy is institutionalized in the Common Core. CCC8015 Generative Artificial Intelligence is a required Year 1 course covering latent spaces, GANs, diffusion models, prompting, fine-tuning, and evaluation of generated content, sitting alongside CCC8011 Critical Thinking — positioning "Critical Thinking 2.0" as the ability to audit AI outputs, detect LLM hallucinations, and critique algorithmic bias.
• Curricula are renewed, not merely augmented, with AI. This includes a new School of Data Science as a "liberal arts hub for STEM," interdisciplinary majors (Creative Media Industries, Animation & Digital Arts, Social Data Science, BA in Philosophy & AI), sunsetting programs with weak five-year employment prospects, and integrating AI into non-STEM faculties — with strong graduate indicators (99.1% employment among those joining the workforce).
• Value systems and cultural heritage are retained as a guardrail. The Department of Philosophy pivots to a "Philosophy of AI" hub using comparative ethics (e.g., Zhuangzi and Daoist conceptions), digital humanities projects preserve heritage, and Service-Learning evolves into "humanitarian technology" (e.g., PureAura, designed for Hong Kong's subdivided flats, which won 14 international awards).

What this means for practice

• Administrators. Choose adoption with access over prohibition: Lingnan has provided campus-wide free access to advanced GenAI tools since August 2023, on the reasoning that bans are futile and that digital fluency should not depend on a student's ability to pay.
• Administrators. Institutionalize AI literacy in the core rather than the elective list: a Generative AI course (CCC8015) is required in Year 1 alongside Critical Thinking (CCC8011), framing the audit of model outputs, hallucinations, and algorithmic bias as critical thinking itself, and treating digital fluency as a liberal arts competency rather than an add-on.
• Instructors. Keep the human at the gate in assessment: in the GenAI Assessment System, AI analyzes submissions against instructor-supplied rubrics but its feedback is routed to the instructor for review and validation before a student sees it, which preserves consistency and timeliness without removing judgment.
• Administrators. Spend infrastructure money on humanistic aims rather than around them: the case pairs a new School of Data Science and interdisciplinary AI majors with a Philosophy of AI hub, digital humanities preservation, and service learning that became humanitarian technology, so that empathy and service leadership stay in the graduate profile next to critical auditing of AI output.
• Administrators. Expect the blueprint to need translation: institutions with limited resources, more centralized or decentralized governance, or stricter data regulation face greater barriers than a well-resourced liberal arts university.

Limitations

• This is a position paper and institutional case analysis, not an outcome study: the author states that the claims remain primarily conceptual and interpretive, supported by descriptive accounts, faculty self-reports, early pilot feedback, and course documentation rather than comprehensive empirical data.
• The account is single-institution and self-reported: it is written by Lingnan's own president about Lingnan's transformation, so the adoption-first policy, the GenAI Assessment System workflow, and the curriculum renewal are described by their architect rather than independently evaluated.
• No learning-outcome, adoption-effect, or equity-impact data are reported; the author states that comprehensive empirical validation awaits future research and that the blueprint is aspirational and forward-looking rather than definitive causal evidence.
• Institution-level figures such as the 99.1% employment rate among graduates entering the workforce, or the 14 international awards for PureAura, describe one university's context and cannot be attributed to the AI strategy.

Connected Concepts

• Higher Education
• Humanities and Social Science Education
• Generative AI
• Change Management
• Curriculum Design
• Human-in-the-Loop

Connected Articles

• Control vs. Agency: Exploring the History of AI in Education — historical "control vs. agency" framing of AI transformation in education (Mishra, Henriksen, Woo & Oster 2025)
• Illuminating complex student realities of artificial intelligence through an entangled pedagogy framework — qualitative entangled-pedagogy analysis of how higher-ed students negotiate generative AI in their lived realities (Fawns et al. 2026)
• Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task — neural evidence of "cognitive debt" from GenAI-assisted essay writing in higher education (Kosmyna et al. 2025)

Citation

AI for education: The digital transformation of a liberal arts institution – implementation at Lingnan University — Qin, S. J. (2026). Computers and Education: Artificial Intelligence, 10, 100592.