Artificial Intelligence and Architecture

Artificial intelligence (AI) has begun to make its mark in the field of architecture, not as a replacement for architects but as a powerful collaborator that extends the creative and analytical capabilities of professionals. In the architectural design process, generative AI systems already make it possible to create multiple conceptual schemes based on initial parameters, optimize spatial configurations, and explore materials through automated simulations. For example, recent studies show how generative models support the conception phase by proposing alternatives for spatial distribution, structures, and envelopes based on criteria of efficiency, sustainability, and response to the physical context.

Beyond conceptual design, AI is integrated into performance analysis: intelligent tools can evaluate energy consumption, natural lighting, ventilation, and thermal comfort, allowing fast feedback to the designer to adjust the proposal. This use also favors sustainable architecture strategies by facilitating informed decisions about orientation, materials, and passive systems. However, for AI to be useful in architecture, it must be human-centered. This means that systems must consider the preferences, customs, needs, and perceptions of real users, not just quantitative optimizations. Otherwise, the resulting designs may lack meaning or sensitivity to human use.

In the professional sphere, recent studies indicate that the adoption of AI in architectural firms is still low, although enthusiasm is growing. According to a report by the American Institute of Architects (AIA), only about 8% of firm leaders report active use of AI, although a larger percentage are considering its future implementation, especially for support in routine tasks and conceptual generation.

The integration of artificial intelligence into the human design process presents significant challenges. One of the main ones is the lack of transparency of many models, known as “black boxes,” which makes it difficult to understand the reasons behind the solutions generated and limits the architect's creative control. In addition, there is a risk of over-reliance on algorithms, which could lead to homogenization of design and a loss of aesthetic diversity. Added to this are relevant ethical and technical issues, such as the authorship of works created with AI assistance, liability for possible errors, and the protection of data used in these systems.

Looking ahead, research is exploring workflows such as Sketch-to-Architecture, where simple sketches are transformed into conceptual plans and 3D models using AI, accelerating guided ideation. Another innovative proposal—the SCAPE system—combines evolutionary algorithms with generative AI to inject creativity and variation into conceptual designs.

In conclusion, AI is emerging as a powerful assistant for architecture: it allows for accelerated iterations, integrated performance analysis, and informed decision-making. However, its real value depends on how it is integrated with human judgment, respect for spatial experience, and the demand for transparency and ethics in its use.

References:

• Li, P., Li, B., & Li, Z. (2024). Sketch-to-Architecture: Generative AI-aided Architectural Design. arXiv.

• Lim, S. L., Bentley, P. J., & Ishikawa, F. (2024). SCAPE: Searching Conceptual Architecture Prompts using Evolution. arXiv.

• The role of artificial intelligence (AI) in architectural design. (2025). Springer.

• Towards human-centered artificial intelligence in architecture. (2023). ScienceDirect / AEC industry.

• Architects are excited about the potential of AI, but concerns abound. (2025). American Institute of Architects (AIA).

• Generative AI Models for Different Steps in Architectural Design. (2024). arXiv.

• AI in architecture trends, opportunities and challenges. (2025). GAF blog.