How Are Leading Architecture Firms Using AI in 2026?
This article is based on “How architects use and will use AI in 2026 and beyond” by Phil Bernstein and Vincent Guerrero of Yale School of Architecture, originally published in RIBA Journal.
Artificial intelligence is rapidly transforming architectural practice, moving beyond image generation into knowledge management, design analysis, workflow automation, and software development. In their article for RIBA Journal, Yale School of Architecture researchers Phil Bernstein and Vincent Guerrero explore how architecture firms are currently using AI and how these technologies may reshape the profession over the next few years.
The authors argue that architecture has traditionally been less about solving clearly defined problems and more about what architectural historian Stanford Anderson called “problem worrying” — an iterative process of exploring, testing, and refining ideas. While AI excels at processing information and generating solutions, the creative and judgment-based aspects of architectural design remain fundamentally human.
Based on discussions with architecture, engineering, and construction (AEC) firms, the article identifies four major areas of AI adoption:
1. Visual Design Workflows
AI-powered tools are increasingly used to generate and modify renderings, create videos, and develop immersive virtual environments. Images are no longer the final product but the starting point for richer design experiences.
2. Practice Knowledge Management
Large language models are helping firms organize, search, and retrieve internal knowledge, making project experience and technical expertise more accessible across organizations.
3. Design Analysis and Optimization
Advanced AI systems can evaluate multiple design options simultaneously, considering factors such as structure, energy performance, cost, and sustainability, enabling data-informed decision-making.
4. Software Creation Through Natural Language
Architects can now create custom applications, workflows, and digital tools using AI-assisted coding platforms, reducing the barrier between design practice and software development.
Despite these advances, the authors highlight a significant challenge: fragmented and inconsistent project data across the AEC industry. Without better standards and interoperability, AI systems will continue to face limitations.
Looking ahead, the article suggests that AI will increasingly automate specific tasks rather than replace architects entirely. As technology evolves, architects may spend less time producing information and more time evaluating options, guiding processes, and making strategic decisions.
The central question remains: can AI support the human qualities that make architecture meaningful, or will automation eventually diminish the creative process itself? For now, the authors conclude that while computers generate possibilities, humans remain responsible for judgment, intention, and design vision.