A survey conducted by Mapei Kft. in July 2026, using an online self-completion questionnaire (CAWI), polled 120 construction professionals to assess how clients' use of artificial intelligence affects collaboration with contractors. According to the results, 73% of respondents have encountered clients who cited AI when questioning or verifying professional recommendations, work quality or pricing.
How often do clients invoke AI?
In the survey, 27% of professionals reported they have not yet met clients who referenced AI. Among the rest, 26% said it happens very rarely, another 26% occasionally, and 20% frequently.
What do clients use AI for?
Respondents said clients most often invoke AI regarding technological solutions (35%). Material choice was mentioned by 31%, and necessary work processes by 25%. AI was cited in pricing discussions by 20%, when explaining causes of defects by 17%, and regarding expected execution time by 16%.
The pattern suggests clients typically use AI before work begins — to search for solutions, materials, processes and prices. The issue arises when such preliminary research hardens into an immutable position.
Professionals' concerns: a chatbot cannot inspect the site
Markovich Béla, CEO of Mapei Kft., emphasized: “A chatbot works from general information. It does not see the condition of a wall, what is under a covering or substrate, whether the layer build-up is correct, or what problems appear during demolition. Without these, it can at best answer for an ideal situation, while in construction the ideal case is the rarest.”
Overall, roughly seven out of ten professionals find it in some degree troubling when a client uses AI to check professional recommendations. Broken down: 29% are not bothered at all, 22% are slightly bothered, 27% find it moderately bothersome, and 23% consider it quite or very bothersome.
Open responses do not indicate blanket technophobia. Some professionals accept that clients consult multiple sources and believe collaboration can be easier with a better-informed client, provided professionals explain the on-site realities. Critically minded respondents object when clients treat a chatbot’s general or incorrect statement as authoritative expert opinion. The two most frequent complaints are that AI does not know site conditions and can provide inaccurate or unverified information.
Several respondents pointed out a problem when clients trust a chatbot more than an experienced contractor: this can lead to unrealistic price expectations, unachievable deadlines, or technological solutions inappropriate for the situation. One respondent summarized the dilemma: AI "would solve everything, but it does not take responsibility for the consequences and emerging defects." This underscores that while advice may come from a machine, the contractor ultimately bears responsibility for execution.
AI is also part of professionals' toolset
Among the respondents, 36% use AI at least occasionally, 27% have tried it once or twice, 13% use it weekly, 7% daily, and 17% have never used it.
Professionals primarily use AI to search for technical information and to perform assistant and administrative tasks. Use cases reported include:
- 44%: searching for information about raw materials, construction technologies and manufacturer recommendations
- 22%: cost calculations, determining material needs, or preparing quotes
- 21%: diagnosing technical problems and troubleshooting
- 11%: learning new technologies and professional development
- 10%: planning work processes
AI also takes on more communication and administrative roles: 18% use it for content creation (social posts, videos), 17% for marketing, and 16% for drafting client emails, offers, work logs, minutes and other administrative documents.
Conclusion: a useful tool, not a replacement
Mapei’s survey shows that most construction professionals accept clients using AI for research and idea gathering. The clear limit is when AI shifts from being an information source to a virtual technical inspector that replaces on-site assessment and professional judgment.
While AI can aid preparation, information gathering and administration, it cannot perceive the actual condition of walls, substrates and layer builds, nor can it assume responsibility for on-site decisions. If a client insists on implementing an AI-recommended solution without weighing professional input and site conditions, the result may be professional shortcomings, faulty or botched workmanship instead of the expected proper execution.
Methodology
The exploratory questionnaire surveyed 120 construction professionals. Data collection took place in July 2026 via an online self-completion questionnaire (CAWI).
