Scaffolding AI in Teaching: Insights from AI for Educators
Akpene Agbo, UX Designer
The recent AI for Educators: Sharing Current Practice, Shaping Future Directions event highlighted how colleagues across Queen Mary are scaffolding the introduction of AI into teaching in structured and supportive ways. Karen Hudson, Innovation & Learning Manager, Queen Mary Academy outlined the university’s new Critical AI Literacy approach, which helps students and staff develop an understanding of how AI works before moving on to responsible, critical use. Dr Lilian Schofield introduced the GenAI Curriculum Alignment Model, which guides Schools in gradually embedding AI into learning outcomes, assessments and activities so that AI becomes part of a coherent learning journey rather than an added complication.
During the QM Question Time panel, contributors including Dr Aisha Abuelmaatti, Professor Chie Adachi, Professor Adrian Armstrong, Ali Elbarky, Dr Paulo Oliva and Professor Alastair Robertson stressed that successful AI adoption depends on transparency and guided practice. Students need to understand why particular tools are used, what ethical expectations apply and how AI can support rather than replace human reasoning. The panel agreed that assessment design, not detection tools, is central to helping students demonstrate their own thinking.
The strongest example of scaffolded integration came from the case study shared by Professors Yue Chen and Michael Chai. Their GenAI-empowered group assessment used a five-stage structure that introduced AI gradually, beginning with guided exploration and progressing to independent use, critical validation and oral assessment. Delivered in both UK and Transnational Education, the project used a real-world technical challenge to help students learn when and how to rely on AI, and when to question it. Students reported higher engagement and increased AI literacy, and staff observed deeper understanding, although the case study also highlighted the need for early AI training and careful management of workloads. Across the event, a clear theme emerged. Scaffolding is essential for helping students navigate new technologies confidently and ethically. By introducing AI step by step, aligning it with meaningful tasks and supporting students in developing critical evaluation skills, colleagues across Queen Mary are creating environments where technology enhances learning and prepares students for an AI enabled future.
Report on CODE Webinar - Setting the scene: How institutions approach online learning and teaching.
Violet Chan, Digital Education Support Officer
In a recent Centre for Online and Distance Education (CODE) webinar, Bobbi Moore PFHEA, Education Strategic Lead at the University of Southampton's Centre for Higher Education Practice (CHEP), addressed one of the most pressing challenges facing higher education today: fragmentation in digital education and how institutions can build meaningful, institution-wide frameworks to address it.
Moore highlighted that digital education in many institutions hasn't developed in a coordinated way. Instead, organisations are grappling with "a combination of legacy systems and newer platforms, which do not integrate well," according to the Times Higher Education (2024) Digital Maturity Index report. This fragmentation creates inefficiencies, limits data-driven decision-making, and ultimately affects the quality of digital learning and student outcomes.
During the webinar, attendees voted on the primary causes of fragmentation in their organisations, with structural barriers and departmental silos emerging as the top concern. Staff development was also raised as a critical issue.
Frameworks as the Solution
Moore positioned well-designed frameworks as enablers that can prompt dialogue, guide design, and clarify what's needed for quality education. To be effective, the frameworks need to be living and iterative documents that integrate seamlessly into existing structures.
Moore shared her experience co-leading a strategic project at University of Southampton to develop guidance on online course design and delivery. The university had no coherent framework, only fragmented materials and handbooks. Through co-design with staff and students, they developed new guidance that was adopted as a standard, reflecting the university's culture and values while providing practical tools for staff reflection.
Five Stages for Developing an Institutional Framework
Moore outlined a practical five-stage approach for developing institutional frameworks:
- Define the strategic gap and set the ambition
- Translate strategy and evidence to practice
- Secure leadership and stakeholder buy-in
- Implement with purpose
- Sustain momentum and evolve
Moore also shared her work from the University of Southampton: Common Framework for Online Education, developed by Helen Carmichael and Bobbi Moore, which provides a model for other institutions seeking to bring coherence to their digital education provision.
You can review the recording on CODE YouTube channel.
Laying the Groundwork for Next-Level Learning
Johnny Lee, Senior Learning Designer
The ALT M25 Autumn Meeting held in the London School of Economics and Political Science (LSE) explores how AI, accessibility, and assessment innovation are reshaping what “next-level learning” can look like in practice.
- Steve Rowett’s session from UCL provided a pedagogical provocation about what “next-level learning” requires in an AI-saturated environment. He argued that if educators only expose students to polished chatbot interfaces, we reduce AI engagement to surface-level prompt consumption — leaving learners positioned as passengers rather than critical operators of socio-technical systems. By demonstrating UCL’s proof-of-concept work, Steve repositioned AI agent-building as a form of computational and critical literacy: a way for students to understand data structures, model behaviour, reasoning chains, and the implications of design decisions. The project is a space for applied inquiry and responsible practice, aligning closely with contemporary calls for deeper AI fluency in higher education.
- Giselle Tadman’s contribution from LSE shifted the focus from tools to the underlying learning principles that define next-level learning. Drawing on an expanded conception of learner development — active, self-directed, reflective, whole-person, and culturally aware — she positioned assessment as where pedagogical values become visible. Giselle emphasised Gradescope’s role in supporting assessment designs that cultivate transfer, metacognition, and inclusivity. Her framing highlighted the importance of transparent criteria, iterative feedback cycles, and environments where students can see how their thinking evolves over time. Technology serves as an enabler for assessment literacy, learning practices, and equitable student experience, offering a foundation for educators to re-evaluate what meaningful assessment looks like in an AI-mediated world.
Artificial Intelligence at Jisc - Prompting for teaching and learning staff
Johnny Lee
In November, Jisc delivered a practical workshop on “Prompting for Teaching and Learning Staff,” offering an accessible introduction to AI literacy and showing how educators can use different types of prompts to enhance planning, assessment design, and day-to-day teaching tasks. The session walked participants through core prompting approaches—such as instructional, conversational, creative, and iterative prompts—supported by concrete examples and live demonstrations. It also highlighted emerging AI capabilities, including image prompting, text-to-speech, file uploads, and evaluating AI-generated outputs through fact checking and critical questioning.
The workshop also focused on helping staff develop practical habits for effective prompting. Facilitators emphasised refining prompts through cycles of iteration, evaluating responses for accuracy and fitness for purpose, and using AI tools to support lesson planning, feedback generation, and inclusive language review.
Integrating Generative AI within curricula - TIRIgogy CPD Series
Jorge Freire, Senior Learning Designer
On 25 November, TIRIgogy hosted Integrating Generative AI Within Curricula, a well-attended session led by Professor Xue Zhou and Dr Lilian Schofield. The workshop cut through the noise around GenAI and focused on an important gap in research and practice: a coherent, programme-level approach to AI literacy and an integration into teaching, learning and assessment approaches. The presenters demonstrated how existing frameworks, Bloom’s taxonomy and UNESCO competencies can be used to map where AI capabilities are currently taught, where gaps exist and how ethics should sit across all levels of cognitive demand. Their case studies from MSc and BSc programmes showed what this looks like in practice—co-creation with students, alignment with assessment, and a realistic view of staff concerns through the TPACK model.
The Bloom-mapping exercise shared is a practical diagnostic tool programmes can adopt, and the framing of ethics as a thread rather than a standalone topic was particularly strong. The resources shared and the session recording are worth exploring, as the shift away from frameworks, principles and analysis towards intentional, curriculum-wide design and teaching practice is a welcome step forward.