AI is reshaping construction, but the biggest transformation may not be about replacing people.
Source: Supplied. Roelof van den Berg, CEO of Gap Infrastructure Corporation (GIC).
As the industry grapples with skills shortages, rising demand and increasingly complex projects, technology is emerging as a tool to help scarce expertise go further. From planning and project management to inspections and paperwork, AI can remove routine pressures while leaving critical decisions in human hands. The challenge now is finding where automation adds value without sacrificing experience, judgement and accountability.
Experience from other industries shows why the human role remains essential. Ford Motor Company leaned too heavily on automated systems, then spent three years bringing back more than 350 veteran engineers and technical specialists to catch defects the technology missed. Similar lessons have emerged in agriculture, mining, ports, and utilities, where automation performs best when skilled people remain able to supervise and intervene.
Construction makes this need especially clear because sites are variable, crowded, and constantly changing. International trials of robotic bricklayers and autonomous heavy machinery show real potential, yet wide adoption remains difficult when equipment must respond to highly unpredictable conditions. The strongest near-term opportunity lies in helping workers and project teams plan, co-ordinate, inspect, and make decisions more effectively.
Construction’s human capital remains its advantage
Plumbers, bricklayers, electricians, roofers, and painters are among the many skilled construction workers AI and robotics will struggle to replicate for the foreseeable future. Most contractors understand that human judgement, creativity, and adaptability are critical site skills that machine learning cannot reproduce, particularly in unstructured environments where conditions change without warning and workers must make decisions in the moment.
The debate too often starts with replacement. I believe the better starting point is to ask which pressures technology can remove from scarce skilled teams and where it can give them clearer information, stronger controls, and more time for the work that requires judgement.
One of the construction sector’s biggest challenges is not job loss to automation. In fact, the sector is hiring more and building at a scale it hasn’t seen in years, having added 130,000 new jobs in the third quarter of last year alone – close to half of all new employment created nationally.
The issue is instead a shortage of workers in key fields. Nearly four in 10 respondents to the latest Critical Skills Survey Report struggled to fill engineering roles in 2025, with only one engineer available for every 3,166 South Africans. The sector also faces a critical artisan shortage, with 22% of companies struggling to fill those positions.
When the Project Management Institute projected this year that there will be a 57% shortfall in construction project professionals across sub-Saharan Africa by 2035, it named building information modelling, digital twins, and AI as the levers most likely to close the efficiency gap.
In this environment, AI and automation become support systems and effort multipliers. The businesses gaining an edge are using technology to redesign work, moving scarce engineers and site teams off routine tasks and onto the decisions that require training, experience, and accountability.
AI is a capability multiplier, not a substitute
AI is already proving useful around construction, particularly in drafting rooms and project offices, where data-rich processes are easier to digitise and scale. On a project, that means AI can help teams plan and sequence work, flag design clashes before they reach site, track progress against programme, catch likely rework early, and move the surrounding paperwork faster, from procurement checks and document trails to payment certificates and project records that would otherwise sit across diaries, spreadsheets, drawings, inboxes, and finance platforms.
One pattern I have seen repeatedly on large projects is that delays often begin in the information surrounding the physical work. An approval sits too long, a drawing is outdated, a record is missing, or a payment certificate cannot move because the supporting documents are scattered across different systems. Technology creates value when it clears these blockages before they slow the people delivering the project.
Outcomes will continue to improve as the technology and the sector’s understanding of it continue to develop. Engineers will spend fewer hours reconciling records by hand, project managers will spot earlier when cost, programme, or documentation begins to drift, and site teams will face fewer hold-ups from missing approvals, duplicated checks, or delayed certification.
For construction companies under pressure to deliver more with scarce skills, AI can empower those doing the work by clearing and strengthening the systems around them. These advances will support the construction and infrastructure sector’s biggest advantage: it’s people.
