From disruption to design: Developing an adaptive AI in education framework for higher education

Feature article

By Professor Heidi Le Sueur

ICMS Deputy Vice Chancellor (Learning and Teaching)

Introduction

When ChatGPT was launched in November 2022, it reached one-hundred-million users within two months of its release, becoming the fastest-growing consumer application in history[1]. This did not just signal a technological breakthrough, it created both complex regulatory challenges and opportunities in higher education globally[2], [3]. Assessments could be completed in minutes, feedback could be automated, and the boundary between human and machine-generated work became increasingly blurred. Almost overnight, higher education providers were confronted with a critical question: how can we preserve learning, integrity, and fairness in an AI-enabled world?

Since then, generative artificial intelligence (Gen AI) has disrupted higher education at great speed, challenging established approaches to learning, teaching, and assessment. The growing difficulty of distinguishing between human and AI-generated work has heightened concerns about academic integrity and assessment validity, as highlighted by Bittle and El-Gayar[4]. Notwithstanding, Gen AI has also created many opportunities in higher education. As Johnston et al.[5] have identified, drawing on survey and focus group interview data at the University of Liverpool, students are more supportive of engaging with Gen AI tools to assist with their learning, especially for tasks such as editing, summarising, and idea generation as opposed to trusting Gen AI to in writing up an entire assignment. Students highlighted the need for clear policies and guidelines to avoid its misuse and enable ethical and responsible engagement that can support student learning and to prepare them for an AI-enabled work environment.

As the discourse around the use of Gen AI evolves, institutions have been shifting from prohibition towards guided integration of Gen AI within ethical and pedagogical frameworks. Similarly, the International College of Management, Sydney (ICMS) developed its Artificial Intelligence in Education (AIED) Framework[6] and accompanying guidelines[7] between 2023 and 2025, with regular review cycles embedded throughout their development and implementation.

Rather than creating a fixed policy response, the institution adopted an iterative, evidence-informed approach designed to evolve alongside changing technologies, regulatory expectations, and stakeholder needs. In practice, this meant recognising that the institutional response would remain provisional rather than final, as developments in AI were occurring more rapidly than traditional policy and review cycles could accommodate.

The challenge is ambiguity in a changing landscape

As institutions move from prohibition towards integration, questions remain about how Gen AI can be incorporated while maintaining educational quality, assessment validity, and academic integrity[6]. Increasingly, higher education institutions recognise that banning AI is neither practical nor desirable, particularly as its use by students and staff continues to expand.

However, integrating Gen AI into learning, teaching, and assessments presents challenges. While AI can offer benefits to enhance personalised learning, student engagement, and timely feedback, it also raises concerns around academic integrity, ethical use, data privacy, and the risk of over-reliance when it is not carefully embedded within sound pedagogical design[8].

The literature highlights the value of an iterative approach, where policies and practices are continually reviewed and adapted as the technology evolves[6]. Many of the challenges sit at the human level. Staff capability, confidence, and readiness to engage with these tools remain uneven, influenced by technical constraints and varying levels of AI literacy[8]. Recognising this situation, the move towards institutional frameworks and practical guidelines, such as those implemented at ICMS, reflects a broader sector response.

Designing a framework in a time of flux

Based on this context, ICMS set out to develop an approach that balanced clarity with flexibility. The AIED Framework was designed not as a static policy document, but as a structured and adaptive framework capable of guiding decision-making across teaching, learning, and assessment, underpinned by the principle of ‘human-in-the-centre’ as articulated in the 2025 version of the framework[6]. This human-centric principle means that Gen AI can enhance learning, but decisions about judgment, ethics, and assessment remain with educators and students.

Revisions of the framework introduced structured approaches to assessment design that align acceptable Gen AI use with learning outcomes, supporting students in developing critical thinking and ethical judgment. This required moving beyond a binary approach of allowing or prohibiting Gen AI towards more nuanced models of engagement through different ‘assessment tracks’, distinguishing between assessments designed to assure learning and assessments designed to develop AI capability. This reflects a shift seen across institutions, where assessment is being rethought to deliberately incorporate, rather than avoid, AI technologies[9].

Transparency has also emerged as a central theme during the development and iterative revisions of the AIED framework. What became clear during consultations was that uncertainty was often a greater source of concern than AI itself. Staff and students were generally willing to engage with the technology but wanted greater clarity regarding expectations and boundaries. For example, students consistently highlighted the need for clarity regarding AI-supported feedback from lecturers on their assessments.

In the guidelines, the use of generative AI in feedback is described as a considered extension of existing teaching practice, supporting editing, enhancement, and guidance, while remaining clearly bounded by requirements for transparency and human oversight. At the same time, the use of Gen AI raised important points around consistency in application and data protection, which means that integration depends not only on formal policy settings, but on capability development for students and academics to align expectations and practice. These findings informed the development of elements within the ‘Use of AI in Assessment Guideline’, particular those focused on clarity, accountability, and purposeful integration of AI to provide practical guidance for the implementation of the framework.

Learning through iteration and refining a framework

The AIED Framework is defined by its iterative approach to development and implementation consisting of four stages with integrated stakeholder feedback and reviews. This ensures, rather than being introduced as a top-down completed solution, it evolves through cycles of stakeholder engagement and evidence-based improvements. The four iterative stages of development, comprising (1) policy and guidelines, (2) institutional learning, (3) student and community engagement and (4) improvements and transformation, facilitates a staged implementation which is responsive to feedback collected in each stage[6], [10].

Throughout the initial development phase, focus groups, pilot projects, surveys, and institutional monitoring data provided insights into how the use of Gen AI and institutional expectations are experienced in practice. For ICMS, the data did more than highlight challenges, it pointed to what needed to change and brought forward some practical ideas. For instance, lecturer feedback highlighting administrative complexity led to the integration of Gen AI enhanced processes into some tasks such as rubric design, aligning assessment briefs with learning outcomes and improving consistency. Similarly, student feedback revealed a lack of confidence in using AI tools responsibly, leading to the inclusion of a compulsory AI literacy module in the students’ learning experience at the institution.

The ongoing refinements have helped in keeping the framework relevant as both the technology and regulatory expectations continued to change. More broadly, it suggests that AI-related policy needs to be viewed as an evolving system rather than a fixed endpoint.

Insights emerging from practice

The implementation and review of the AIED Framework have generated several key insights with broader sector relevance. One of the strongest messages from both students and staff has been the need for clarity. When expectations around AI use are explicitly defined, students are more likely to engage confidently and ethically, while ambiguity often undermines learning outcomes.

Importantly, the way assessments are designed has emerged to be one of the critical factors. Traditional assessment approaches, often focused on evaluation of final outputs, are increasingly exposed to Gen AI-generated responses. In contrast, process-based and authentic assessments with in-class demonstration of applied skills provide stronger evidence of student learning, aligning with recommendations in the AIED framework[6]. As a result, assessment design remains one of the primary mechanisms for assuring learning.

Framework implementation also depends on staff confidence and capability development, with research emphasising that academic staff AI literacy and skills are critical aspects for effective adoption[11]. As reflected in the framework review process at ICMS, policy and guidelines alone are not sufficient, and targeted professional development and education is essential to support application and reduce variability across subjects.

Yet, equity considerations remain a critical concern. Differences in access to Gen AI tools and AI literacy may create disparities in student outcomes, reflecting broader ethical challenges[12]. Addressing this requires not only clear policy guidance but also institutional investment in accessible, secure AI technologies.

Implications for the higher education sector

Looking across the implementation process, the ICMS experience shows that responding to AI is not only a technical or regulatory challenge, but a fundamentally educational one. Institutions must rethink how learning, teaching, and assessment are designed in an AI-enabled environment, moving towards approaches that are flexible, adaptive, and evidence-informed.

Emerging regulatory expectations also point towards more structured and accountable approaches to AI integration in higher education⁹. A consistent lesson throughout implementation was that governance and policy cannot sit apart from pedagogy and teaching practice. Policies that operate in isolation from teaching realities are unlikely to succeed. Instead, frameworks must be embedded within everyday educational practice, supported by staff capability development and institutional processes.

Perhaps most importantly, the policy review and implementation process must remain grounded in stakeholder experience and feedback to remain relevant, practical, and responsive to real-world challenges.

Conclusion

The integration of Gen AI into higher education represents one of the most significant transformations in recent decades. AI technologies present challenges for higher education, but they also make institutions to rethink how to design learning and innovate. The AIED Framework development experience demonstrates that effective responses to Gen AI are built through structured, evidence-informed, and iterative processes that integrate governance, pedagogy, and stakeholder engagement. This approach reflects broader sector shifts towards adaptive and responsive policy design.

As Gen AI continues to evolve, so too must the frameworks that guide its use. Institutions that prioritise adaptability, transparency, and human-centred design will be better positioned not merely to respond to technological change, but to shape its educational value.

Acknowledgement

The author declares no conflict of interest and does not have any financial disclosures.

To cite this article:

Le Sueur, H (2026, July 8). From disruption to design: Developing an adaptive AI in education framework for higher education. Scholarly Impact. International College of Management, Sydney. https://www.icms.edu.au/scholarly-impact/learning-and-teaching/ai-education-framework/

References

[1] Hu, K. (2023, February 2). ChatGPT sets record for fastest-growing user base – analyst note. Reuters. https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/

[2] UNESCO. (2023). Artificial intelligence in education. United Nations Educational, Scientific and Cultural Organization. https://www.unesco.org/en/digital-education/artificial-intelligence

[3] Tertiary Education Quality and Standards Agency (TEQSA). (2025). Gen AI knowledge hub. https://www.teqsa.gov.au/guides-resources/higher-education-good-practice-hub/gen-ai-knowledge-hub

[4] Bittle, K., & El-Gayar, O. (2025). Generative AI and academic integrity in higher education: A systematic review and research agenda. Information, 16(4), 296. https://doi.org/10.3390/info16040296

[5] Johnston, H., Wells, R. F., Shanks, E. M., Boey, T., & Parsons, B. N. (2024). Student perspectives on the use of generative artificial intelligence technologies in higher education. International Journal for Educational Integrity, 20, 2. https://doi.org/10.1007/s40979-024-00149-4

[6] International College of Management Sydney (ICMS). (2025a). Artificial Intelligence in Education (AIED) Framework. https://policies.icms.edu.au/artificial-intelligence-in-education-aied-framework/

[7] International College of Management Sydney (ICMS). (2025b). Use of AI in Assessment Guidelines. https://policies.icms.edu.au/use-of-artificial-intelligence-ai-in-assessment-guidelines/

[8] Garzon, J., Patino, E. & Marulanda, C. (2025). Systematic review of artificial intelligence in education: Trends, benefits, and challenges. Multimodal Technologies and Interaction, 9, 84. https://doi.org/10.3390/mti9080084

[9] An, Y., Yu, J. H., & James, S. (2025). Investigating higher education institutions’ guidelines and policies regarding the use of generative AI in teaching and learning. International Journal of Educational Technology in Higher Education, 22, 10.  https://doi.org/10.1186/s41239-025-00507-3

[10] Tertiary Education Quality and Standards Agency (TEQSA). (2024). Gen AI strategies for Australian higher education: Emerging practice toolkit. https://www.teqsa.gov.au/guides-resources/resources/corporate-publications/gen-ai-strategies-australian-higher-education-emerging-practice

[11] Nikolic, S., Wentworth, I., Sheridan, L., Moss, S., Duursma, E., Jones, R. A., Ros, M., & Middleton, R. (2024). A systematic literature review of attitudes, intentions and behaviours of teaching academics pertaining to AI and generative AI (GenAI) in higher education: An analysis of GenAI adoption using the UTAUT framework. Australasian Journal of Educational Technology, 40(6), 56–75. https://doi.org/10.14742/ajet.9643

[12] Marín, Y. R., Caro, O. C., Carrasco Rituay, A. M., Guimac Llanos, K. A., Pérez, D. T., Sánchez Bardales, E., Alva Tuesta, J. N., & Chávez Santos, R. (2025). Ethical challenges associated with the use of artificial intelligence in university education. Journal of Academic Ethics, 23, 2443–2467. https://doi.org/10.1007/s10805-025-09660-w