The adoption and use of AI in the built environment sector is increasing, with around a quarter of companies surveyed by Pinsent Masons and industry leaders already taking active steps to adapt their business models to significant future disruption. However, feedback also shows that the industry is unwilling to adopt and utilize both general and tailored AI solutions, preventing it from leveraging the full range of productivity improvements that AI can bring.
Pinsent Masons, together with Bentley Systems, Mott MacDonald and Turner & Townsend, sought the views of a wide range of built environment organizations globally between February and June 2025. Respondents included architectural and engineering design consultants, contractors, project and cost managers, technology providers, and facility owners and operators.
Of the organizations surveyed, 24% identified AI-driven change as a major issue their companies are currently grappling with and said they have already taken active steps to adapt their business models in anticipation of significant future disruption. Although 40% of respondents expect AI to significantly impact their business models in the coming years, only 20% said AI is currently strategically important to them or that they have plans to use it more broadly in their organizations.
When it comes to the use of AI, respondents most commonly cited uses for automating document-related processes. Real-time monitoring of operational data to predict when maintenance is required or to optimize and improve the performance and resilience of infrastructure and building assets; and optimization of construction productivity. Where companies are already investing in AI, there is already a strong focus on using AI systems to optimize and improve the productivity of design and construction processes.
When asked about their specific AI use to date, 36% of those surveyed said they used generic AI systems such as Copilot. While only 15% of respondents said they had used AI selectively or tested applications and had not yet integrated the use of AI into their operations, 26% said their companies were already using and testing AI systems for specific use cases.
The use of agentic AI systems – which use “static” AI agents and other tools and resources to build additional skills such as reasoning, planning and self-assessment – does not appear to be widespread in the industry yet. However, only a small proportion of respondents said that they were not actively thinking about introducing or using AI at all.
Ian Laing, infrastructure and property expert at Pinsent Masons, said the answers should encourage organizations across the construction sector to think about where and how AI can be used most effectively. “Optimizing and improving the productivity of design and construction processes is a current focus,” he said. “This includes generative design, complex simulation, digital twins and rapid solution analysis that can help optimize design and engineering solutions.”
“Other focal points include the automation of document processes as well as cost estimates and forecasts.”
A significant proportion – two-thirds – of architectural and engineering designers surveyed are either already taking active steps to adapt or foresee significant future disruption to their current operational business models.
While there has already been progress in the use of AI-powered automation in sectors such as mining, the construction industry, particularly construction contracting, has historically been much slower to transition and adopt new technologies. This is confirmed by the survey data, which shows noticeable reluctance and skepticism among contractors towards AI. Only 12% of contractors are actively taking steps to adapt their current business models to future disruptions. While 35% of contractors expect AI to have a significant impact on their business model in the next three years, over 50% currently expect AI to have limited or no impact.
Graham Robinson, infrastructure expert at Pinsent Masons, said AI has the potential to disrupt even key construction tasks sooner than many contractors might think. “While practitioners are skeptical about on-site automation in the short term, the long-term potential remains high and contractors who ignore this potential could fall victim to the rapid evolution of AI,” he warned. “Contractors could now control AI and robotics in safe, incremental ways – such as with autonomous drones for site surveys or AI for security monitoring – so as not to fall behind as on-site automation accelerates.”
Of the 86 survey respondents, 75% held management or board level roles in the construction sector, with contributions coming primarily from the UK, North America and Australia, with some responses from Europe, Asia, the Middle East and Africa.
According to the experts, the results of the survey on governance and risk are all the more convincing. While 60% of organizations surveyed said they had some type of AI policy, 20% said they had a general organization-wide AI policy in place. Only 6% of companies said they have specific policies for individual departments that set guidelines for the use of AI, governance, ethical implications, security measures and related aspects, and plan to update the policies. Another 12% said they had no department-specific or organization-wide AI policies at all.
About 37% of companies said they currently have limited or no project controls to manage risk when using AI for design and construction. Such gaps could expose organizations in the built environment to further unwanted risks and stresses, Robinson said. “Without proper governance and project controls, AI failures could create security risks or legal liabilities,” he said. “Validation of AI models, peer review of AI-powered designs, and clear policies and contractual terms around AI use will help ensure there are appropriate checks and balances in place.” He said regulators may also need to update guidelines to cover AI-assisted engineering.
Looking forward, more than two-thirds of respondents said they expect more than half of their projects to make significant use of AI in the design and construction of infrastructure and buildings within the next three years. A total of 29% of respondents expect similar progress in construction, suggesting that there is still significant room for expanding AI use cases in this market segment.
According to Oxford Economics, global construction spending is expected to reach $10 trillion in 2025. With the disruptive and transformative impact of AI expected to increase in pace and momentum over the next few years, Laing says the built environment sector must be prepared to adopt, leverage and invest in tailored AI solutions to maximize productivity gains in an increasingly unpredictable business climate. “Significant investment will be required to transform an isolated and highly fragmented industry into a more integrated, digital, data-driven and modernized alliance of diverse capabilities,” he said.
Robinson said a mindset and culture shift, coupled with significant investment in digital skills and training, will also be crucial to helping the sector keep up with the dynamics of AI. “Assuming there will be a majority or higher value projects in three years that make significant use of AI, this will require a significant shift in the conservatism and culture of the construction industry, and the actual pace of adoption could be slower if the challenges are not addressed,” he said. “The obvious bridge between readiness and maturity needs to be addressed. Challenges such as talent deployment and the need for much greater data standardization are important considerations.”