The latest UAL AI Network meeting brought together colleagues from across the university to explore how artificial intelligence is being integrated into teaching, creative practice and research. Rather than focusing on AI as a technological solution, the discussion centred on a more important question: how can we help students and staff engage with AI critically, ethically and creatively?
Three presentations highlighted different approaches, from graphic design education and cultural research to academic inquiry, revealing both the opportunities and limitations of AI in creative education.
Creating Space for Students to Experiment
The first presentation, led by Lynn Kiang (BA Graphic Design, Camberwell) and supported by Chris Follows, explored a live project developed in partnership with Outernet. Students were challenged to create immersive experiences for one of London’s largest digital public spaces while exploring the role of AI throughout the design process.
Importantly, AI was not positioned as a tool for replacing creativity. Instead, students were encouraged to use it selectively for research, project management, coding, translation, data interpretation and experimentation. The aim was to help students develop their own informed position on AI rather than prescribing how it should be used.
One of the most interesting findings was that many students chose not to use AI for final image-making. While they found value in AI-assisted research and organisation, they often wanted to retain ownership of the creative outcomes themselves. Students frequently moved back and forth between AI tools and manual design processes, discovering that meaningful results required substantial iteration, prompting and critical judgement.
The project also raised important conversations around ethics, transparency and documentation. Students were asked to record their AI use, including models, workflows and key prompts. This encouraged a more reflective approach and helped students develop greater AI literacy alongside their design skills.
What Happens When AI Cannot See the Full Picture?
The second presentation, from Leigh Odimah at London College of Fashion, examined a very different challenge: the gaps and biases within AI training data.
Drawing on a personal archive of Black British magazines, music publications and visual culture dating back to the late 1980s, Leigh explored how AI tools interpret materials that are largely absent from mainstream digital archives. Many of these publications are difficult to access and are underrepresented in the datasets used to train contemporary AI systems.
Using tools including DALL·E, Midjourney, Firefly and Runway, Leigh attempted to recreate images and cultural artefacts from Black British sound-system culture. While the AI tools could often reproduce visual features such as clothing, hairstyles and graphic styles, they struggled to understand the deeper cultural context behind the images.
In several examples, the tools introduced cultural assumptions and stereotypes not present in the original material. The results highlighted how AI often fills gaps in knowledge with associations drawn from dominant datasets, revealing biases that can distort representation.
The project offers a powerful reminder that AI-generated outputs are shaped by what exists within training data—and by what is missing. For educators and students alike, this underlines the importance of questioning AI outputs rather than accepting them at face value.
Becoming “Creative Cyborgs”
The final presentation came from Ray Grewal, who shared research from his MA Academic Practice dissertation and subsequent publications exploring AI’s role in academic thinking and creative practice.
Ray described his extensive use of AI during the research and writing process, using multiple tools as tutors, research assistants, translators and data-processing aids. He argued that AI can be highly effective for navigating information, summarising complex materials and supporting routine academic tasks.
However, his key argument was that AI cannot replace the deeper thinking processes involved in learning and research. He introduced the concept of “analytic rumination” — a process involving reading, questioning, synthesising, writing, reflecting and revising ideas over time. These are the activities through which genuine understanding emerges.
While AI can generate fluent text, Ray argued that it cannot engage in this reflective and developmental process. As a result, students risk mistaking polished outputs for meaningful thought if they rely too heavily on AI-generated content.
Looking Ahead
A common thread connected all three presentations: AI works best when it becomes a catalyst for critical thinking rather than a shortcut around it. Whether students were defining their own ethical boundaries, exposing cultural biases in datasets, or reflecting on the nature of human creativity, the most valuable learning emerged through questioning, discussion and experimentation.
As UAL continues to develop its approach to AI, these projects demonstrate the importance of creating spaces where students and staff can engage openly with the technology, explore its potential, challenge its limitations and shape its future use within creative education.
Leave a Reply