AI in Architecture – How is AI Impacting Building Design?
A recent report published by the Architects’ Journal and research carried out by the RIBA revealed that 41 percent of architects were already using AI, even on the occasional project. The research showed that more than two thirds of employees working at large practices are currently using AI. More notably, 15 percent said they regularly use generative AI tools. The role of AI in architecture is expanding, bringing both opportunities and challenges.
Over the past two decades, the development of new tools has been transforming how architecture practices operate. From hand-drawing concepts, to the creation of drawings in CAD, to more recent 3D parametric modelling and BIM, architects have consistently been innovators in adopting digital tools. The rapid development of artificial intelligence and architecture technologies creates new possibilities—redefining architectural design, workflow, and efficiency. While AI has the potential to help architects solve complex design challenges, critical questions remain. It’s essential that the integration of AI tools in everyday practice enhances efficiency and creativity, rather than diminishing creative thinking.
Examining the impact of artificial intelligence
AI within the architecture industry refers to the application of computational technologies and algorithms to assist architects throughout various stages of the design process. Common areas where AI for architectural design is currently used include generative design, project management, and BIM. 3D parametric modelling using BIM has been embraced by architects since its introduction in the early 2000s, bringing greater collaboration between technical and creative disciplines. However, despite BIM’s impact, it still struggles to streamline workflows across the entire design process. This is where AI has the potential to step in—simplifying concept design, integrating environmental and sustainability concepts, and optimizing AI building design performance.
Moreover, using artificial intelligence for architectural design allows for iterative testing and design validation in ways that traditional processes cannot match. Machine learning models trained on historical performance data can suggest design optimisations with speed and accuracy, potentially reducing both cost and waste. As projects grow in complexity and as timelines become tighter, these efficiencies can make the difference between a project succeeding or stalling.
How is AI being used for architectural design?
The data from RIBA revealed that AI is primarily used during the concept stages of projects—preparing ideas and early-stage documentation. Tools such as ChatGPT for text generation and Midjourney for image generation are popular among architectural practices. Generative AI architecture tools can efficiently create visuals, enabling architects to explore a broad array of design iterations. AI used by architectural practices tends to fall into four main categories:
1. Design Ideation
The most common use of AI in an architecture practice is for design ideation and visualisation. Tools like Photoshop’s Generative Fill allow users to create compelling visuals with simple text prompts. This is especially beneficial for smaller practices, where specialist rendering services may no longer need to be outsourced. In this way, AI helps democratise access to sophisticated presentation techniques that were previously limited to larger firms.
2. Concept Designs
AI can assist architects in applying stylistic iterations to early-stage models, enabling quick adjustments to materials, textures, and visual aesthetics. Rather than rebuilding a model from scratch, designers can now rely on AI tools to produce alternative forms and materials almost instantaneously, allowing for broader creative exploration.
3. Text Generation
Survey results show that architecture practices use AI tools such as ChatGPT for writing and editing reports, fee proposals, planning applications, and administrative documents—saving time and effort. This alleviates time-consuming tasks and allows architects to focus on design-specific activities. Some practices are even developing custom AI tools that mirror their writing style and technical vocabulary.
4. Practice Archiving
Larger practices are beginning to use machine learning to train AI on their previous design data and libraries. This can not only streamline workflows and boost efficiency, but also help avoid copyright or authorship concerns by keeping the process in-house. An AI trained on a firm’s body of work may suggest solutions that align with its brand, values, and aesthetic principles.
Overcoming industry obstacles with artificial intelligence
With growing pressures on the architecture industry—such as climate change, urban density, and economic demands—the creativity of the profession is increasingly tested. Architects must design high-performing buildings that enhance quality of life, often within tight constraints. AI can support this challenge by enabling outcome-based design. By automating routine tasks, using AI for architectural design, can streamline workflows and allow architects more time to focus on creative problem solving. It can also help arrive at solutions more quickly and efficiently.
AI can integrate environmental and contextual data into designs, offering real-time analytics on factors such as daylight, wind, microclimate, noise, and operational energy. In the context of the climate crisis, this is critical. Beyond providing data, AI enables digital testing and simulations to find optimal solutions. Designers can simulate building performance under different conditions long before breaking ground, improving sustainability outcomes and reducing energy consumption.
The challenges of AI use in architecture
While AI offers significant benefits to the architectural profession, many practitioners remain cautious. According to a survey by Architects’ Journal, 36 percent of respondents viewed AI as a threat to the profession. Although AI can reduce monotonous tasks, there are concerns that it cannot replicate core architectural responsibilities.
Tasks that traditionally require skill, time, and labour can now be replicated instantly with AI. Despite this, the technology is still in its early stages and far from replacing human creativity and design expertise. Authorship and intellectual property are also areas of concern, as the integration of AI raises questions about copyright and ethical practice.
Rather than replacing architects, AI can serve as an assistant—supporting designers while they retain control of the creative process. Architects understand their clients’ needs, from aesthetic to regulatory requirements. Ultimately, it is the architect who is responsible for creating better buildings and homes. By combining the architect’s knowledge and experience with AI’s computational power, we can unlock new design possibilities. Architects will remain irreplaceable, with more time to focus on solving complex challenges.
Throughout history, architects have adapted to technological advances while maintaining a deep understanding of human needs and cultural values. AI cannot replicate the storytelling, sensitivity to site and context, or the nuanced decision-making that architects bring to the table. While generative AI can produce impressive visuals, there remains a gap between conceptual imagery and buildable, regulation-compliant architecture. Under current Building Regulations, architects are still responsible for ensuring final designs meet material, structural, and legal requirements. Misinformation or inaccuracies in AI outputs may outweigh the time saved—highlighting that the technology is not yet at a point where AI can fully transform architectural work.
While AI has transformative potential in the architecture industry, its adoption is not without barriers. A major challenge is the upfront investment required, which can be particularly difficult for smaller firms. Learning to use AI tools effectively also demands time and training. However, this investment could allow architects—especially in small practices—to offload administrative work and refocus on core design challenges.
There is also a cultural dimension to this technological shift. Architectural identity is often rooted in locality, heritage, and the narrative of place. Many fear that over-reliance on AI could lead to homogeneity in design, where projects lose their distinctiveness in favour of efficiency. Striking a balance between innovation and cultural preservation will be crucial as the technology matures.
The future of AI in architecture
The RIBA survey explored future outlooks on AI adoption. It found that 54 percent of respondents believe their practices will adopt AI within two years, while 25 percent disagreed. Most respondents agreed that AI would be used to support environmental sustainability analysis and improve workflow efficiency. While many are optimistic, others maintain that AI will always have limitations—especially in addressing the cultural, historical, and social dimensions of design, as well as the intricate spatial, structural, and regulatory demands. AI also lacks the ability to respond to subjective client needs and preferences.
As AI continues to evolve—much like CAD and BIM before it—it should be seen as a tool to support architects, not replace them. By embracing this technology with care and critical thinking, the profession has the opportunity to advance into a more efficient, inclusive, and imaginative future.
Our final thoughts
At Rodic Davidson Architects, we see the use of AI in architectural design as an exciting and evolving tool with the potential to support and enhance our design process. As a small practice, we are particularly interested in how this technology can increase efficiency and allow us to focus more on creative, high-value tasks. We remain open to integrating AI where it adds value, particularly in early-stage design and visualisation. However, while we recognise its potential, AI cannot replicate the judgement, intuition, and contextual understanding that define our role as architects. The technology may assist us—but it cannot replace us.