Dive Brief:
- Artificial intelligence is changing construction teams more often than shrinking them, according to a report from Berkeley Research Group.
- Among North American respondents, 46% changed team structures and 38% retrained employees, while 18% added headcount in certain areas and 11% reduced it because of AI, per the report.
- Over two-thirds of North American respondents invested at least $250,000 in AI over the previous two years, including 37% that spent $1 million or more, the survey found. Despite the widespread adoption and spend, 15% said they’d fully integrated the technology across projects and operations.
Dive Insight:
BRG surveyed 109 executives involved in AI implementation in July 2026, including 58 builders and 51 owners. Most findings reflect 65 North American respondents. Surveyed builders were mainly small- to mid-sized general contractors and construction managers; owners were primarily larger developers and real estate investment trusts.
The workforce changes uncovered extend to recruiting as 29% of respondents created new types of roles and 20% changed the skills they seek when hiring. Another 31% redeployed staff to different work, and 51% changed their reliance on outside vendors, consultants or contractors.
Richard Fultineer, BRG’s managing director and head of global construction, said the shift centers on how companies handle project information.
“In recruiting, it is critical to focus on team members who will be able to manage and progress AI-type applications in the future,” he said. “The immediate impact is on information-intensive work like cost estimating, procurement, design review, schedule oversight, and contract and compliance review.”
Eight out of 10 surveyed North American builders use AI to help evaluate and plan proposed projects, per the report. Fultineer said AI tools can also review construction drawings to spot errors or missing details that could lead to costly corrections.
“Another example is an estimating model’s ability to draw on hundreds of completed, similar projects to calculate material, labor and equipment costs,” he said.
Scaling those tools presents challenges beyond buying software. Forty percent cited a lack of skilled talent or expertise as a challenge, while 43% identified liability for AI-generated errors as a concern restricting use.
Closing that skills gap requires training tailored to each phase of construction, Fultineer said.
“Much of this occurs in-house but requires constant usage to become the norm,” he said. “Encouragement at the senior levels of companies is necessary to push for efficient and continued use.”
Construction’s adoption depth was similar to BRG’s surveys of retail, energy and financial institutions. AI was active in most core functions for 42% of construction respondents, compared with 48% in retail and 54% in financial institutions However, workforce retraining and hiring changes cannot be compared because the construction survey was the first to ask about them.
Construction also had the lowest reported policy coverage among the four sectors for human oversight and accountability at 42%, and accuracy and reliability at 34%. Fultineer attributed that gap partly to banking and energy’s stronger regulatory pressure to formalize oversight.
Contractors should track estimating accuracy, schedule performance, rework and change orders to evaluate results, Fultineer said.
“Overall, it is a hands-on practice,” he said. “They should practice on ongoing projects with proper consideration and continuing quality control to check and avoid errors.”
