Skip to main content
Queen Mary Academy
  1. Queen Mary Academy
  2. Educators
  3. Innovation and Scholarship
  4. Innovative Pedagogies
  5. Centre for Excellence in AI in Education
  6. Blog
  7. Items

From TikTok Workouts to Curriculum Design: Why AI Competence Needs a Programme-Level Approach

Generative AI is rapidly transforming how engineering students learn and solve problems. While global frameworks now define what AI competence should include, translating these ideas into coherent degree programmes remains a challenge. Professor Yue Chen explores why AI competence cannot be developed through isolated modules and introduces AICoDE, a framework designed to support programme-level curriculum design in engineering education.

Published:
Profile photo of Professor Yue Chen

When AI competence development is left to chance 

Imagine trying to improve your fitness by following random exercise videos on TikTok. 

One day you try a short abdominal workout. The next day a high-intensity leg workout. A week later, you discover a stretching trend that promises instant results. Each exercise might be useful on its own. But without a structured training plan, the results quickly become uneven. Some muscle groups are overtrained, others neglected, and meaningful progress becomes difficult to achieve. 

In many ways, students’ development of AI competence risks following a similar pattern. 

As generative AI tools rapidly enter higher education, educators and modules are responding in different ways. Some modules actively integrate AI into learning activities and projects. Others restrict its use. Some simply avoid addressing it. 

From a programme perspective, this can create a fragmented learning experience. Students’ exposure to AI depends on the modules they take and the preferences of individual module organisers. They may encounter intensive AI use in one semester and none in the next. 

Instead of building competence progressively, their experience with AI becomes a series of isolated encounters with AI. 

Yet the ability to work effectively with AI is becoming an essential, particularly for future engineers and technology professionals. Engineers increasingly work in environments where AI systems assist with modelling, design exploration, predictive analysis, and decision support. Developing the ability to collaborate with such systems requires more than knowing how to use tools. It also requires critical judgement, ethical awareness, and domain-specific understanding. 

Like physical training, developing this competence requires intentional progression: foundational understanding, guided practice, and increasingly complex applications over time. 

Why programme-level thinking matters 

In recent years, a growing number of organisations and researchers have proposed frameworks describing what AI competence should include. 

For example, the UNESCO AI Competency Framework for Students highlights the importance of combining technical understanding with ethical awareness and human-centred perspectives in AI education (UNESCO, 2024). Similarly, research on AI literacy emphasises that effective AI competence involves not only technical knowledge but also the ability to critically evaluate AI outputs and collaborate productively with AI systems (Chiu, 2025). 

These frameworks provide important guidance for what students should learn about AI. 

However, translating these competencies into coherent curriculum design is not straightforward, especially in engineering programmes, where curricula must already integrate mathematics, theoretical foundations, laboratory practice, and industry-relevant technical skills. 

Engineering education typically develops skill progressively; students move from learning fundamental principles to applying them in increasingly complex systems and design projects. AI competence should follow a similar trajectory. Early modules may focus on understanding AI capabilities and limitations, while later modules may incorporate AI into domain-specific tasks such as system modelling, data analysis, optimisation, or intelligent network management. 

Without programme-level coordination, these opportunities remain disconnected.  

This raises an important question for educators and programme leaders: 

How can universities translate AI competence frameworks into curriculum structures that support students’ development across their entire learning journey? 

Introducing AICoDE 

To address this challenge, colleagues in the School of Electronic Engineering and Computer Science at Queen Mary University of London have been developing AICoDE (AI Competence Development in Education) (Chen et al., 2026). 

The core idea behind AICoDE is that AI competence should be scaffolded intentionally across the curriculum, rather than appearing sporadically in individual modules. 

The model combines two complementary perspectives. 

AI competence development 

The first dimension focuses on how AI competence evolves across the student learning journey and develops through three tiers (Chen et al., 2025): 

  • Foundational literacy – understanding the principles, capabilities, limitations, and ethical implications of generative AI 
  • Technical proficiency – applying AI tools in discipline-specific contexts such as coding, data analysis, modelling, and system design 
  • Domain-specific problem solving – integrating AI critically and creatively to address complex engineering challenges and create transformative solutions 

This developmental approach aligns with broader AI competence frameworks while focusing specifically on how competence progresses across a degree programme. 

AI-aware assessment design 

The second dimension addresses assessment design in AI-rich learning environments. 

As AI tools increasingly generate text, code, and analytical outputs, traditional assessments may no longer reliably capture students’ reasoning processes. To address this challenge, AICoDE incorporates the 3A Assessment Framework (Alfadhl et al., 2025). 

The framework defines three broad assessment types based on levels of AI engagement: 

  • AIM – AI for Initial Mapping, where AI may support pre-assessment learning, such as brainstorming, ideation, revision or scoping, but AI use is not permitted during the assessment itself. This ensures that core human competence remains visible and assessable. 
  • ACT – AI for Cognitive Tasks, where AI supports structured reasoning, analysis and simulation. Students must verify and interpret AI-generated outputs through reflection and critical evaluation, ensuring that human oversight and interpretive skills remain central.  
  • APT – AI for Problem Solving and Transformation, where AI is used in open-ended, design-focused or investigative tasks. Students must synthesis ideas, integrate AI effectively into their workflow, and ensure responsible use of AI to address complex real-world problems.   

By linking competence development with assessment design, AICoDE provides a dual-axis developmental model, where competence tiers (vertical axis) intersect with assessment types (horizontal axis). Each module is situated within a single competence tier, but may contain multiple assessment components mapped to different 3A categories, reflecting the authentic complexity of curriculum and assessment design. 

From individual modules to curriculum insight 

AICoDE is designed not only as a conceptual model but also as a tool for programme-level curriculum review and enhancement. The implementation involves four steps: 

  1. Curriculum review – mapping modules to competence tiers 
  2. Pedagogy evaluation – examining how learning activities support AI competence 
  3. Assessment audit – analysing how assessments engage with AI 
  4. Impact analysis – visualising patterns through a programme-level dashboard 

The resulting dashboard provides programme leaders with a clear overview of competence progression and assessment practices across the curriculum. It helps identify structural gaps, such as underrepresentation of higher-tier competence modules or over-reliance on particular type of assessments. 

Such insights allow educators to move beyond isolated innovations and towards evidence-informed curriculum enhancement. 

Preparing students for an AI-enabled future 

While AICoDE has been developed within an engineering education context, the underlying principle is not limited to engineering. Across many disciplines, from the sciences and social sciences to the humanities, students will increasingly encounter AI as part of their learning and future professional practice. 

Ensuring that learners develop the ability to use AI critically, responsibly, and effectively therefore requires similar questions to be asked across curricula: How should AI competence develop over time? How can learning activities support that development? And how should assessment evolve to reflect meaningful human–AI collaboration? 

Returning to the gym metaphor: meaningful progress rarely comes from following random exercise videos. Instead, improvement happens through a structured training programme that builds capability step by step. 

Developing AI competence in higher education should follow the same principle. It needs to be designed intentionally across the curriculum, rather than left to chance. 

The full paper is available here: 

Chen, Y., Chai, K. K., Shu, C., & Alfadhl, Y. (2026). AICoDE: AI Competence Development in Engineering Education. 2026 IEEE Global Engineering Education Conference (EDUCON). IEEE, 2026. 

References and Further Reading 

Alfadhl, Y., Chen, Y., Chai, M., and Zhou, X. (2025). Integrating artificial intelligence (AI) into assessments: The 3A framework for higher education. Educational Developments, 26(4), pp.24–26. SEDA. Available at https://www.seda.ac.uk/wp-content/uploads/2026/01/EdDevs-26.4_DEC_2025_FINAL.pdf  

Chen, Y., et al. (2025). A GenAI Competence Framework for Engineering Curriculum Enhancement in Higher Education. Intelligent Technologies in Education. Advanced Online Publication. Available at https://doi.org/10.53761/ITED/1.2 

Chiu, T. K. (2025). AI literacy and competency: definitions, frameworks, development and future research directions. Interactive Learning Environments, 33(5), 3225-3229. Available at https://www.unesco.org/en/articles/ai-competency-framework-students 

Miao, F., & Shiohira, K. (2024). AI competency framework for students. UNESCO Publishing. Available at https://doi.org/10.54675/JKJB9835 

Professor Yue Chen 

Director of Scholarship

https://www.qmul.ac.uk/eecs/people/profiles/chenyue.html 

 

 

Back to top