A recent survey by Kahua indicates that while nearly all construction firms have adopted AI, very few are seeing actual returns. CEO Scott Unger told BNN Bloomberg that this gap exists because companies often rely on generic tools rather than embedding AI into their core operational workflows .
The 8 percent of firms finding actual value
The disparity between usage and utility in the construction sector is stark. According to the Kahua survey, approximately 92 percent of construction firms report using some form of artificial intelligence, yet only 8 percent claim to see a clear, measurable benefit from that investment. This suggests that for the vast majority of the industry, AI is currently a novelty or a superficial addition rather than a value-driver.
This pattern echoes a broader trend across several legacy industries where "AI theater"—the adoption of tools to appear innovative—outpaces the actual restructuring of business processes. As the Kahua report suggests, the small minority of firms reaping rewards are those that have moved beyond surface-level experimentation to integrate AI directly into their business ecosystems.
Why native integration beats generic ChatGPT wrappers
A primary reason for the low ROI is the method of deployment.. Scott Unger, the chief executive officer of Kahua, noted in his interview with BNN Bloomberg that many firms simply use AI as a productivity layer, such as hooking up ChatGPT to a field engineer's tablet. While these generic models can assist with basic tasks, they lack the specific project context required for high-stakes construction decisions.
In contrast, native AI integration—where the technology is built directly into the software managing the sequence of construction activities—allows for real-time decisions aligned with specific project parameters. when AI is embedded in a core tool , it can automate complex tasks like submittal reviews or asset label interpretation with far greater accuracy than a standalone chatbot.
The risk of unverified data in construction management
Despite the potential for automation, the construction industry cannot yet remove the human from the loop. Scott Unger emphasized that AI is "not a magic bullet" and warned that the risk of misinformation is high if construction managers accept automatically generated data without cross-checking. Because a mistake in a construction sequence can lead to masssive cost overruns or safety hazards , human oversight remains non-negotiable.
The necessity for trained personnel to verify AI outputs highlights a critical bottleneck: the industry's lack of specialized AI literacy. The value of the tool is currently capped by the ability of the human professional to judge whether the AI's output is a factual insight or a hallucination.
Canada's hesitation compared to U.S. AI confidence
The survey also revealed a geographical divide in how technology is perceived. While AI usage is high across North America,Canadian firms displayed significantly more caution and hesitancy than their U.S. counterparts. Scott Unger attributed this disparity to a lack of familiarity and a deficiency in education regarding the technology's reliable application.
This raises several specific questions that the source does not fully answer: What specific "industry-specific training modules" are required to bridge this gap? Furthermore, the report does not specify what the 8 percent of successful firms define as a "measurable benefit"—whether this refers to reduced labor costs, shorter project timelines, or decreased waste.. Without these benchmarks, other firms may struggle to replicate that success.
Comments 0