Beyond the render: embedding AI tools into architectural education

How student-led research is shaping the future of visualisation workflows.

Artificial Intelligence (AI) is changing the way technologists and architects design, visualise, and share ideas. From quick concept sketches to near-photorealistic renders, new tools are reshaping workflows across the built environment. For graduates preparing to step into practice, architectural education must adapt quickly to ensure the next generation is confident in using these technologies.

This student-staff research project at Nottingham Trent University was conducted as a desktop study, focused specifically on how AI tools might be integrated into architectural visualisation within education. Using workshop data, literature analysis, and a critical review of teaching practices, the project explored how AI could support creativity, accelerate workflows, and improve student engagement, while also raising questions about authorship, ethics, and accuracy. By blending academic research with practical insight, the study begins to outline strategies for embedding AI into curriculum design, helping students navigate the opportunities and challenges of an AI-assisted future.

This project was a collaboration between students, lecturers, and researchers at Nottingham Trent University, aimed at exploring the educational potential of AI in architectural visualisation. Conducted as a desktop study, all evaluations were carried out through literature analysis, digital testing, and structured workshops, rather than live projects or studio teaching.

The study had three main goals: (1) to assess the creative and technical potential of AI in an educational context across RIBA Stages 0 to 4, (2) to enhance student coursework using a carefully selected range of tools, and (3) to develop practical guidance for lecturers to support ethical and effective integration of AI into workflows. Tools were shortlisted from academic and industry sources reviewed during the literature phase, with a further filter to ensure alignment with NTU’s IT policies. All selected platforms were web-based, allowing easier implementation in institutional settings. Aligned with the RIBA Plan of Work and learning outcomes, the project highlights how student-led research can drive innovation in architectural education while responding to rapid industry change.

Our methodology focused on evaluating AI within architectural education. We began by reviewing 15 sources, including academic publications and industry articles, to understand the current landscape of AI in architecture. This informed our tool selection criteria, which considered RIBA stage relevance, educational accessibility, visual output quality, and alignment with NTU’s browser-based IT policies. Ten AI platforms were taken forward to a student workshop for testing, using standardised prompts and baseline visuals for direct comparison. Ten students from the Architectural Technology programme, including Master’s students from the Digital Architecture and Construction programme, evaluated each tool based on usability, output quality, and how well results aligned with original design intent. A parallel staff workshop gathered academic perspectives on AI integration, addressing themes such as authorship, assessment, prompt writing, and its wider educational role. Participants included staff from the Architectural Technology, Architecture and Interior Architecture programmes at NTU.

This combination of literature review, structured tool evaluation, and collaborative feedback allowed the research to capture both student experience and teaching priorities within a controlled academic setting. Students consistently highlighted the advantages of AI-assisted visualisation during design development. The ability to iterate quickly meant ideas could be tested and refined in less time, improving both creative flow and visual communication. Many students reported increased confidence in presenting their work and appreciated the ability to explore design variations with ease. However, issues also emerged. Some AI tools introduced unintended changes to elements such as furniture or structure, which led to a loss of design accuracy. Output quality also varied depending on how prompts were written. One student reflected that “tools are only as good as the prompt,” reinforcing the importance of developing prompt-writing skills as part of digital literacy. These insights highlight that while AI can enhance design workflows, its value depends on thoughtful use and a strong grasp on design intent.

Lecturers recognised its potential but raised concerns around academic integrity and authorship. How can originality be assessed if much of the visual output is AI-generated? The answer, perhaps, lies in shifting the focus of assessment towards process and intent rather than just final visuals – a key principle within the draft lecturers’ guidelines. By valuing how students develop and articulate their ideas, AI becomes a support tool, not a shortcut.

Key findings include: (1) Tool performance varied depending on the design stage, with some platforms better suited to early conceptual development and others more effective for refined visualisation. This reinforced the need for strategic integration of AI tools in line with RIBA Stage intentions. (2) Prompt-writing proved critical. Students who developed clear and spatially descriptive inputs consistently produced stronger results, supporting the introduction of prompt-writing workshops within design modules. (3) Questions around ethics and authorship emerged throughout. Issues of originality, intellectual property, and student ownership require clear academic guidance and structured assessment frameworks. Ultimately, AI did not replace design thinking, it supported it. Students remained the authors of their work, with AI acting as a creative support that encouraged iteration and improved visual communication, though careful consideration of tool selection and clear authorship must remain a central part of the process.

A key outcome of the project was the development of draft Lecturers’ guidelines. Rather than setting strict rules, these were designed to prompt discussion and adapt over time in line with developments in teaching methods and curriculum delivery. Their purpose is to support the responsible and effective integration of AI into architectural education while encouraging students to reflect critically on its role. The guidelines highlight four focus areas: (1) Delivering prompt-writing workshops to help students develop clarity, creativity, and visual literacy. (2) Embedding ethics-based teaching around authorship, academic integrity, and originality. (3) Encouraging strategic AI use at early design stages for idea generation, while prioritising manual methods in later phases. (4) Framing assessment to value process and design intent as much as the final visual outcome.

These guidelines offer a starting point for ongoing discussion around the role of AI in architectural education. They are intended to be revisited regularly as tools develop, and new challenges arise. Crucially, they reinforce the principle that AI should support, not replace, the core design skills that underpin architectural technology education.

AI tools are evolving faster than most areas of design technology, and some explored during this project may already be outdated. This rapid pace highlights the need for adaptable teaching, with ongoing research embedded into curriculum delivery. Future work could include: (1) Expanding tool evaluations to reflect emerging capabilities, (2) Exploring AI’s role in areas like environmental analysis or visual communication, and (3) Strengthening collaboration across academic teams to improve teaching practice. By treating the curriculum as a flexible framework, architectural education can stay responsive to technological change while equipping students with the critical skills needed in an AI-enhanced design environment.

This project gave us the opportunity to combine research with practical testing, using AI to generate meaningful insights into architectural education. I would like to thank Dr Nacer Bezai, Dr Moulay Chalal and the TILT team for their guidance and support. Funding from TILT enabled access to premium AI tools and delivery of structured workshops. Special thanks also to Jessica Hakes for her valuable contribution throughout the early stages of the project, including the literature review and collaboration in the delivery of workshops. We hope this work encourages further student-led inquiry and sparks continued reflection on the role of AI in architectural learning.

AI is moving quickly, often faster than education can adapt. This raises a new question for architectural schools and beyond: how might the strategies explored in this project be adapted across other design disciplines, and what role should ongoing research play in shaping AI integration across the wider university curriculum?

This article appears in the AT Journal issue 156 Winter 2025 as "Beyond the render: embedding AI tools into architectural education" written by Thomas M Job, Nottingham Trent University.

--CIAT