Industry

AI in our industry – where are we at?

1 September 2026

8 minutes to read

Tangible machinery and subtle technological adaptations are two sides of the spectrum of AI use in Kiwi construction to take a look at as we assess the rapidly shifting sands in this innovative landscape.

WHAT YOU NEED TO KNOW:

New Zealand's construction sector is increasingly adopting AI, from physical robots on-site to everyday digital tools to increase efficiency.
Southbase Construction's Dunedin Hospital project saw a semi-autonomous drilling robot introduced on site and a New Zealand-first robotic layout printer.
Generative AI is reshaping tasks like drafting, contract and cost checking, with much of this ‘shadow AI’ use happening without company oversight.
AI-driven market insight is helping reveal which regions, sectors and clients are generating live work, opening opportunities for more contractors to bid.
Looking ahead, we can expect more robotics on site, predictive project-risk tools, formalised company AI policies, and sector-wide coordination through bodies like the AEC Working Group.

The Kiwi construction sector stands alongside every other industry in its evolving use of AI to increase efficiencies, lower costs and drive innovation. Cost escalations, labour shortages and financial pressures make more streamlined ways of operating attractive. With the availability of AI growing across many strands of the sector, its use is understandably increasing.

From physical applications such as robots drilling holes to algorithms that flag cost blowouts before they happen, AI is becoming more a part of how New Zealand builds. Here's a look at where the technology stands, mid-2026, and where we might expect it to go next.

Design and planning smarter before the first sod is turned

Building Information Modelling (BIM) has been standard practice on larger New Zealand projects for years, and AI is now layered on top of this. It is used to generate design options, check models for clashes between structural and services elements, and run feasibility comparisons that used to take a design team days of work.

The Architecture, Engineering and Construction Working Group was formed under the AI Forum New Zealand, following the 2024 AI Blueprint workshops and has been coordinating much of this shift at an industry level. The group's stated goal is to drive AI adoption across a sector that contributes over $17.6 billion to New Zealand's GDP, or 6.3 percent of the total economy, with workstreams covering everything from workforce development and spatial planning to climate resilience and Te Ao Māori engagement.

Robots on the tools

Perhaps the most visible sign of AI's arrival on New Zealand construction sites is physical robots. A standout example so far is Southbase Construction's work on the New Dunedin Hospital Outpatients Project. Southbase used the project to bring in the Hilti Jaibot, a semi-autonomous drilling robot roughly the size of a large quad bike, believed to be only the third such piece of technology in Australasia to be used at the time. Guided by the project's BIM model, the machine is driven into position by an operator before switching to autonomous mode to drill mounting holes within a set radius of the structure. This is work that would traditionally mean a tradesperson working at height with a drill for hours at a time.

Southbase's BIM manager explained the team had to develop the model further to handle a more complex drilling situation, factoring in numerous brackets, templates, and specific hole widths and depths.

"The safety case is clear: removing the hazard of working at height altogether rather than just managing it.”

The same project also delivered a New Zealand first on the layout side, trialling the HP SitePrint robot. This machine reads a coordinated CAD file exported from the BIM model and autonomously marks out walls, penetrations and fixtures directly onto the slab, complete with sensors to detect unmapped obstacles on site. Southbase's team has described BIM-enabled robotics as part of a broader shift they expect to change New Zealand construction workflows across multiple trades, not just their own.

The quiet, everyday AI

Away from the robotics headlines, a quieter, more widespread shift is happening on laptops and phones across the industry. Generative AI tools are being used for drafting communications, summarising specifications, checking contracts, running cost estimates, and building automated workflows that used to require a developer.

Structural engineers are a good example. At an Engineering New Zealand webinar earlier this year, one automation specialist demonstrated how tools built on platforms like n8n and Cursor can now automate workflows, streamline business processes and support rapid development of practical internal tools for engineering firms. This is task-specific automation that doesn't need a big software budget or an IT department.

This grassroots adoption has a governance downside, though. New Zealand IT advisory firm NSP has flagged a growing pattern of "shadow AI" in construction firms, where staff from estimators to project managers to site supervisors adopt AI tools individually to meet real deadlines, without company oversight of what data is being fed into them.

The firm points to Deloitte's 2026 State of AI in the Enterprise findings that worker access to AI rose by 50 percent in 2025 alone, yet only one in five companies has a mature governance model to manage that access. It notes the problem is amplified in construction specifically because the workforce is distributed across sites, offices and projects, making it harder for management to track than in a single-office business environment. For business owners, that's less an argument against AI and more a signal that a basic usage policy is now overdue on most sites.

Reading the market and the pipeline

AI is also being applied above the level of an individual project, to the flow of work coming into a business. Sector data platforms are using AI-driven analysis of national project pipelines to help contractors decide which opportunities to chase in a tighter market, such as with Hubexo's Construction Outlook for 2026 drawn from national pipeline data, a sentiment survey and industry interviews. Tools that can flag which regions, sectors or clients are actually generating live work give smaller firms some of the market intelligence that just the large contractors with dedicated business development teams used to be privy to.

Where New Zealand is at now

Global research backs up what's happening locally. The Autodesk and Deloitte State of Digital Adoption in the Construction Industry report found 94 percent of Asia-Pacific construction and engineering businesses now have plans to integrate AI and machine learning. This is understandable with New Zealand's challenges of supply chain disruption, rising costs and labour shortages driving the appetite for AI-driven efficiency gains.

New Zealand's relatively small, tightly networked construction sector means genuine world-first trials, like the Jaibot and SitePrint technology in Dunedin, can happen on a single hospital project rather than huge flagship jobs of multinational contractors. Local industry bodies including the Building Institute Aotearoa and Engineering New Zealand are actively running briefings and case-study sessions through 2026 specifically to help firms figure out what’s useful in 2026, and how companies should decide what to adopt and what to avoid.

Where next?

Here are a few likely directions from where the industry stands today:

More robotics on more trades. If the Dunedin Hospital results hold up, layout printing and drilling robots can be expected to move from a headline-grabbing trial to a standard tender inclusion on larger commercial and institutional jobs. This is likely to become prevalent particularly where repetitive, precise, at-height work is involved.
Predictive project control. AI models that flag cost and programme risk early by comparing a live project against historical ones are likely to move from pilot to mainstream, especially as firms operate with thinner margins for error.
Governance catching up with adoption. We can expect more firms to formalise AI use policies, rather than leaving staff to find their own tools.
Sector-wide coordination. Bodies like the AEC Working Group are working on roadmaps, workforce upskilling and Te Ao Māori-informed approaches to AI adoption, for a more coordinated rollout than the ad hoc, firm-by-firm adoption seen so far.
AI being used as a retention tool, not just an efficiency one. Skilled people, their knowledge and experience, will likely become even more valuable assets in an AI environment. Valuable, skilled people are likely to assess and value the way a company uses AI and how this is managed may become an important factor for employees and contractors when they’re choosing who to work for.

Where we’re at as an industry doesn’t appear to be on the precipice of AI running wild to build New Zealand houses unsupervised. More that AI is taking on the repetitive, risky parts of the job and giving experienced tradespeople and project teams more space for the craft, supervision and judgement calls that still need a human on site.