How firms can direct AI instead of reacting to it

In the Bay Area, artificial intelligence is no longer part of an abstract conversation about the future. It is part of daily life: Dinner arrives via delivery robots, driverless cars move through city streets, and the companies shaping the next generation of AI technologies are often only a short drive from where we practice architecture.  

Working in this environment creates a particular kind of pressure. Even if you are not naturally inclined toward technology—which is the case with me—you quickly realize that ignoring it is not an option. When technological change is visible everywhere, it becomes difficult to assume the architectural profession will remain untouched.

The conversations I hear among architects about AI tend to fall into two camps: quiet anxiety or enthusiastic hype. Both responses miss the opportunity for a more productive question. Instead of asking whether AI will change architectural practice, we might ask how architects can guide its influence on our design process and workflows.

Architects have encountered moments like this before. When CAD replaced hand drafting, many worried that speed would erode craft. In some cases, it did. But it also expanded what architects could produce with smaller teams and fewer resources. Over time, the profession absorbed the tool and redirected it toward its own purposes. Artificial intelligence may represent a similar inflection point.

Architects are trained editors. We gather information, filter possibilities, and refine ideas until they become architecture. In that sense, AI is less a replacement for our work and more a system that responds to our direction.

In my own small residential studio in California, I have begun treating AI as a flexible tool: part research assistant, part design collaborator, and part operational support system. Used intentionally, it can expand the creative and analytical capacity of small firms without replacing professional judgment.

Below are several ways AI is beginning to influence how we work.

Expanding research in a more complex regulatory landscape

Residential architects increasingly operate within dense layers of regulation, evolving housing legislation, and local interpretations of state policies. In California, for example, recent housing laws have created new opportunities for lot splits and additional units—but they have also introduced a web of technical constraints that architects must understand early in the design process.

AI has proven useful as a preliminary research tool in these situations. It can help synthesize legislation, surface relevant municipal interpretations, and organize large volumes of information quickly. This does not replace professional verification, but it accelerates the early phase of understanding the parameters that shape a project. For small firms without dedicated researchers, this efficiency allows staff to spend more time on design rather than administrative interpretation.

Supporting iteration on small and constrained sites

Much residential work takes place on sites that are constrained by topography, setbacks, and neighboring structures. A small shift in massing or orientation can dramatically affect daylight, privacy, and spatial quality. 

AI-assisted tools can help explore variations of these spatial ideas quickly during early design phases. Rather than spending hours modeling each possibility, we can test concepts rapidly and evaluate which directions deserve deeper development.

This does not produce finished architecture. It produces a broader field of ideas that the architect can then refine through traditional design methods. In small firms with limited staff, expanding iteration without expanding hours can meaningfully strengthen the design process.

Deepening creative research for interiors and materials

Interior architecture often draws from a wide range of references: art history, craft traditions, regional materials, and architectural precedents. AI has become a surprisingly effective research companion in this area. In our studio, we sometimes use it to trace references to particular craft traditions, explore the history of materials, or identify artists and makers whose work aligns with the atmosphere we are trying to create.

For example, when developing interior specifications, AI can bring forth lesser-known artisans or historical references that enrich the design narrative. These discoveries then guide deeper research and sourcing. The architect still curates and evaluates what is meaningful. AI simply accelerates the discovery process.

Strengthening the operational side of small practice

Running a small firm means balancing design work with the realities of operating a business. Proposals, marketing materials, project descriptions, and client communication all require time that could otherwise be spent on design.

AI tools can assist with drafting proposals, organizing research, refining written content, and preparing materials for communication with clients or collaborators. While these tasks may seem secondary to design, they are essential to sustaining a small studio. Reducing the time required for administrative work allows architects to focus more energy on design leadership and client relationships.

Architects as editors of technology

The long-term trajectory of AI in architecture remains uncertain. It is clear, however, that tools evolve according to how professionals choose to use them.

Architects have always been editors of complex systems, balancing technical requirements, human needs, and aesthetic judgment. Working with AI requires a similar mindset. It is less about automation and more about direction: defining the questions, refining the prompts, and selecting what is worth developing further.

If small firms approach AI in this way, it becomes less of a disruptive force and more of a trainable extension of practice. It can assist with research, support early iteration, and strengthen the operational resilience of studios that often work with limited resources.

Technological change has always shaped architectural practice. What remains constant is the role of the architect: interpreting complexity and translating it into spaces that serve people and place. Rather than waiting to see how AI transforms architectural practice, we have an opportunity to shape how we use the technology within the profession.

Mahya Salehi, AIA, is the founder and principal of Mahya Salehi Studio Architecture + Interiors.