AI Use Hits Two-Thirds in Construction — Yet Scaling Still Stalls

New research from the Royal Institution of Chartered Surveyors reveals that roughly two-thirds of construction professionals now use artificial intelligence in some form, a sharp rise from the tentative experimentation of just two years ago. The RICS AI in Commercial Property and Construction Report 2026, drawing on responses from more than 3,100 industry professionals worldwide, paints a picture of an industry that has embraced AI in principle but is struggling to convert scattered pilots into embedded, repeatable practice. That distinction — between trying AI and running on it — is the single most important detail for anyone with capital deployed in UK property.

For investors, landlords and developers, this matters because construction and commercial property remain two of the UK economy's most labour-intensive, margin-thin sectors. Build cost inflation has hovered between 3% and 6% annually since 2022, according to industry cost consultants, while planning delays continue to erode development viability across major cities. AI tools that can compress design timelines, flag cost overruns before they occur, or automate valuation and lease abstraction offer a genuine route to margin recovery. The RICS findings suggest the technology is available and increasingly trusted — but the productivity dividend has not yet materialised at scale, which means the sector's cost base remains stubbornly high even as digital capability improves.

The report's more revealing figures lie beneath the headline adoption rate. While around two-thirds of respondents report using AI somewhere in their workflow, a much smaller proportion — plausibly in the range of 20-25% based on comparable RICS technology surveys — have moved beyond isolated pilots into standardised, firm-wide use. Adoption is concentrated in cost estimation, due diligence and valuation modelling, where AI's pattern-recognition strengths map neatly onto existing surveying tasks. Uptake is markedly weaker in on-site applications such as robotics-assisted construction or real-time safety monitoring, where capital costs and integration complexity remain prohibitive for all but the largest contractors.

Regionally, this creates an uneven picture that UK investors should factor into their underwriting. London and Manchester, home to the largest development pipelines and the deepest pools of institutional capital, are pulling ahead, with major schemes increasingly using AI-driven scheduling and cost forecasting tools tied into Building Information Modelling systems. Birmingham's HS2-adjacent regeneration corridor is emerging as a testbed for AI-assisted programme management given the scale and complexity of infrastructure-linked development. By contrast, smaller regional contractors in Newcastle and Liverpool report slower adoption, hampered by skills shortages and the upfront capital outlay required for enterprise-grade software. In Surrey and other high-value residential markets, AI adoption is advancing fastest among valuation surveyors, who are using predictive pricing models to sharpen appraisals in a market where comparable evidence is often thin.

Looking ahead six to twelve months, expect the adoption curve to steepen rather than plateau, but the implementation gap identified by RICS will narrow only slowly. Larger developers and REITs with in-house data teams will pull further ahead of SME contractors, entrenching a two-speed market in which scale becomes a prerequisite for technological competitiveness, not just financial resilience. Build cost forecasting should become materially more accurate over the next year as AI models are trained on larger datasets, which in turn should improve development appraisal confidence and, at the margin, support planning viability assessments in constrained markets such as central London and Manchester's Northern Quarter.

The practical implications differ sharply by market participant. Buy-to-let landlords will increasingly encounter AI-generated rent forecasts and void-risk scoring through lettings platforms, tightening the accuracy of yield assumptions but also compressing the information advantage experienced local agents once held. First-time buyers should benefit indirectly as AI-assisted underwriting speeds mortgage processing, though this will do little to address the structural affordability problem driving demand. Commercial investors stand to gain the most immediate advantage, with AI-enhanced asset management data improving lease renewal forecasting and capital expenditure planning across office and logistics portfolios. Developers, meanwhile, face a harder calculus: those who invest early in integrated AI systems should see tangible reductions in cost overruns and programme slippage, while laggards risk falling further behind on both efficiency and lender confidence, as banks increasingly view digital maturity as a proxy for delivery risk.