While artificial intelligence is gradually establishing itself as a major lever for economic transformation, the Building and Public Works sector is moving forward at a measured pace. Less than 10% of business leaders today say they use AI solutions.
The study carried out by Plein Sens on behalf of the Construction Industry Observatory draws a nuanced observation: if use remains limited, interest is growing and the cases of application are diversifying.
Carried out in the first half of 2025, the survey is based on a questionnaire sent to more than 600 managers, supplemented by 28 in-depth interviews. “We did this first questionnaire with 600 leaders in the sector and we supplemented it with more in-depth interviews with business leaders, also with startups that offer solutions based on AI and with training organizations,” explains Olivier Mériaux, Director of Studies at Plein Sens
Adoption still restricted
AI therefore remains marginal in construction companies, with less than 10% of managers reporting using it. More than a third of executives (36%) plan to adopt these technologies in the coming years. The intention is more marked in medium and large companies, half of which anticipate future use, compared to 30% in VSEs.
🚨 SMARTJOBS
- ECOLE POLYTECHNIQUE – Director/Deputy Director of International Relations (F/M)
- CLAROTY — Sales Development Representative
- CURE51 — Data Scientist (Internship)
- FRACTTAL — Account manager (France)
- BRICKSAI — Founding Growth Manager
👉 Find all our offers on the DECODE MEDIA Jobboard
📩 Are you recruiting and want to strengthen your employer brand? Discover our partner offers
This differentiation is largely explained by the structure of the sector. “We have 95% of companies in this sector which have fewer than 9 employees. » This fragmentation mechanically weighs on the capacity for investment and experimentation.
AI enters through the tertiary door
Current uses mainly concern support functions. “Very clearly, the entire study shows that AI is entering the sector through the tertiary sector. » Writing documents, synthesis, analysis of technical briefs, simplification of administrative processes: AI is first used for informational tasks. The core business, on the other hand, remains little impacted. “We are dealing with activities that sometimes lend themselves very little to automation. »
The data barrier
The question of data appears central. “For there to be AI, AI application, you need data. » However, the sector produces data that is often heterogeneous, poorly structured, with uneven archiving practices. The consequence is: “Without quality data, there is no quality AI.” » This reality partly explains the slow diffusion of advanced uses, particularly in terms of design or decision-making automation.
Robotization: the limits of the field
Asked about the robotization of construction sites, Olivier Mériaux is cautious: “There is nothing worse than a construction site environment from the point of view of complexity for robotization systems. » Noise, dust, hazards, unstable surfaces: the construction site remains an environment that is difficult to standardize. The demonstrators exist, but massive deployment seems distant.
ROI and risk culture
Construction remains a pragmatic sector. “We are in a universe, in a culture which will promote (…) reliability, security. » In this environment, AI is perceived as “a black box”, which nourishes a natural reserve. The managers interviewed expect tangible proof: “They want to actually see a concrete, measured, demonstrated benefit. » Some have tested solutions without being convinced. “We have companies that are a little lost in this,” faced with an abundant supply that is sometimes difficult to evaluate.
A primarily human issue
For Olivier Mériaux, the transformation will not only be technological: “We see that like other sectors, construction is not escaping the wave of AI and although its use is limited, its progress is irreversible. What this study reveals is that before being technological, the issue is human and organizational. The sector must collectively build the capacity to make the link between innovation and professions and to support teams over time. »
A diffusion in stages
The study proposes to follow the evolution of uses over time, theartificial intelligence in construction is therefore neither a spectacular breakthrough nor a standstill. It advances according to a gradual trajectory, structured by data, training and organization.
If the technological project is open, it is part of the long term of a sector where transformation can only be carried out to the extent of its operational constraints.
