Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Good foundations: The ROI of AI in the architectural design phase

Understanding the challenges in this stage is key to appreciating how AI can elevate the design process.

Good foundations: The ROI of AI in the architectural design phase

[Nognapas / Adobe Stock]

Profile photo of Sandeep Ahuja

Co-founder & CEO · cove

Construction isn’t getting cheaper, and demand for space isn’t getting smaller. Architecture 2030 projects the world will add 2.6 trillion square feet of new floor space between 2020 and 2060. Not only do we need more buildings, but we need ones that are efficient, cost-effective, and support occupant well-being.

The construction industry is woefully inefficient. A 2015 KPMG report found only 25% of construction projects stayed within 10% of their original deadlines. Research has found that design change is one of the top reasons for cost overrun.

How can we improve the architectural design process to eliminate inefficiencies, identify cost savings, and create better projects? The answer is AI. 

CREATING A BETTER BUILDING PROCESS

It’s not enough to aim for better buildings alone. We need a better building process. Breakthroughs in AI offer powerful solutions—but before AI can revolutionize the process, it must first confront architectural design, the foundation of every successful project.

This phase is characterized by creativity and compromise, where architects strike a balance between aesthetics and function, stakeholder expectations and regulations, and innovation and cost. Understanding the challenges in this stage is key to appreciating how AI can elevate the design process.

THE OBSTACLES OF THE ARCHITECTURAL DESIGN PHASE

1. Predesign

The predesign phase is meticulous. Architects must create a design that fits project goals, stays within budget, and follows zoning laws and building codes. Projects involving specialized industries such as medicine or chemical storage add further complications.

AI can expedite this process by analyzing a specific site’s zoning laws and environmental regulations and indicating potential site-building challenges, such as slope gradient, flood risk, building height restrictions, etc. AI can also research these concurrently, saving more time.

2. Schematic Design

Next is schematic design, where the project’s layout and design begin to take shape. With AI, architects can create numerous iterations of a design in record time, allowing stakeholders to provide feedback and ask for alterations without slowing the process. They can also simulate and test aspects such as energy efficiency and spatial layout. AI-powered cost estimation solutions, which draw from real-time material inventory databases, allow architects to integrate financial considerations earlier in the design process.

3. Developmental Design

This is the stage where architects dive into the design’s technical details. AI streamlines material specification by enabling rapid creation of multiple design versions with varying material choices and associated cost estimates. Teams can customize estimation tools to align with sustainability targets and material preferences. Equally importantly, AI facilitates quick integration of stakeholder feedback, which speeds design iteration and finalization.

CHALLENGES AND COMPLEXITIES OF AI IN ARCHITECTURE

Despite its benefits, like anything new, AI has its drawbacks and challenges. 

AI’s major advantage is its ability to use past data to generate insights and predictions for future projects. A firm could feed an architectural model five years of work and generate a new design based on that data; however, AI’s accuracy depends on the data it’s given. If the project data is full of errors or incomplete, those flaws will be reflected in the results. Unlike humans, who can question information, AI takes what it is given as fact.

This data objectivity can create bias. While AI models themselves aren’t biased, the human-generated data they’re trained on often is. For example, an early version of an image-generating AI was criticized online because, when prompted to generate an image of a “CEO” or “director,” its output primarily featured white men. Why? Stock photography sites in the model’s training data prominently labeled photos of white men as CEOs.

AI models are not magical solutions. They’re tools, and like any tool, people must know how to use them correctly.

KEY CONSIDERATIONS BEFORE BUILDING WITH AI

Architecture firms must be strategic to incorporate AI into their design process. Simply picking a tool and telling employees to use it is a quick way to waste money. Here are a few tips:

• Define your objectives. What areas do you want to improve? Where are your bottlenecks? This will help you find the right tool for your needs.

• Research integrations. AI is useless if it can’t integrate with your most important systems.

• Include employees. Ownership in the process ensures a better fit for your business’s needs and higher adoption rates.

• Start small. Select one project as a test run; when it’s over, talk with your team to refine the process for the next one.

ARCHITECTURAL BUILDING DESIGN WITH AI ASSISTANCE

In the architectural design process, time is money, and so is predictability. AI presents a strategic advantage in an industry known for razor-thin margins and chronic overruns. By embedding AI into the process, architects can achieve faster approvals, reduce design-related rework, optimize material use, and gain early cost clarity—all while preserving design integrity and meeting increasingly complex code and sustainability requirements.

The future of development lies in building better, and it starts with a better process, assisted by AI.