From Resistance to Acceptance: Overcoming Barriers to Large Language Model Integration in Construction Management

The construction industry increasingly adopts advanced technologies, such as Large Language Models (LLMs), to enhance project management processes. However, several challenges hinder the successful integration of these technologies. This research investigates the potential barriers to LLM adoption in construction project management within the South African construction industry. A quantitative approach was employed, with 66 structured questionnaires collected from construction professionals in Polokwane, Limpopo province. The data were analysed using SPSS, and results were presented using central tendency measures, Kruskal–Wallis tests, and dispersion measures. The findings identified the skills gap as the primary barrier to adoption, followed by high investment costs, misinterpretation of LLM results, and ethical concerns related to AI. Based on these findings, the study recommends the development of AI-focused training programs through partnerships between universities offering construction project management education, tech companies, and construction firms. These programs would help bridge the skills gap and ensure the effective deployment of LLMs. To address high investment costs, the study advocates for government funding initiatives, including grants, low-interest loans, development funds, and competitive funding schemes, to support seamless adoption in the sector. To reduce the misinterpretation of LLM results, the study recommends implementing collaborative educational programs, such as workshops, certifications, and hands-on training, to enhance comprehension of LLM outputs. Finally, ethical concerns can be mitigated by establishing clear ethical guidelines, conducting regular audits, and providing ethical training for developers and users to ensure the responsible use of LLMs in construction.