Artificial Intelligence (AI) is rapidly emerging as one of the most transformative technological developments in history. It is already impacting how we live, work and interact with the world; and it is continuing to develop at an extraordinary pace.
The construction industry, including large scale international projects, is no exception.
In this article, we explore:
- What AI means in the context of international construction projects, with a focus on Generative AI; how it works, how it differs from traditional AI, and why it's poised to transform the industry.
- How AI is currently being used across the construction lifecycle, including design optimisation, quality control, project scheduling, health and safety monitoring, and robotics.
- The types of legal disputes that may arise from AI adoption, such as intellectual property ownership, liability for errors, employment law challenges, and data privacy breaches.
What is AI?
AI, in its broadest sense, is not a new concept. It has been used in various forms for many years – for example, predictive text messaging on smartphones and email spam filters are both powered by AI.
However, most of the current attention relates to a specific type of AI, termed "Generative AI".
In 2022, US-based technology company OpenAI launched ChatGPT, a first-of-a-kind chatbot capable of generating "original" content. It is this form of AI that is so revolutionary and is poised to impact our lives so significantly. Since the release of ChatGPT, many other forms of Generative AI models have been developed for different means.
Generative AI can be defined as a form of computer model that has been trained on huge subsets of data (for example, large swathes of the public internet, published books, articles and journals) and which can produce original content in response to a user's prompt.
Much like human intelligence, the model learns the patterns and structures in the language in the data subset on which it has been trained (whether written language, images or audio files), understands a user's prompt and generates original content in response with similar characteristics to that data subset.
The data subset on which the model has been trained can be specifically tailored for different purposes. For example, an AI system could be trained on the entire public archives of the Technology and Construction Court with a view to servicing specific queries related to technology and construction disputes presented to the Court in England and Wales.
Some generative AI tools even claim to be able to predict the outcome of legal claims based on the issues, evidence in dispute and the judge that has been allocated to the case based on historic records.
Unless otherwise stated, where this article uses the acronym "AI", it refers to Generative AI.
Real world applications of AI in international construction projects
AI has already been implemented in the construction industry in a variety of different ways to augment the way the industry operates.
Like any new phenomenon, widespread adoption will take some time and some getting used to. Trust will need to be built through use and success. However, widespread adoption to some extent in the construction industry (and almost every other industry) seems inevitable.
AI in design and modelling
AI augmentation of the design process appears to be increasingly commonplace within the industry.
Some AI tools are able to analyse (and generate) thousands of design permutations based on budget, material, environmental and regulatory constraints. There are examples of AI tools generating designs and 3D models based on simple hand-drawn sketches and text prompts.
Saudi State-owned property developer ROSHN recently stated that it is using AI to create 3D models of buildings as part of its efforts to align with the goals of Saudi Vision 2030 for smart and sustainable cities. Designers and architects will be able to harness the power of these types of tools to augment their processes and achieve greater accuracy and productivity.
AI for material planning and take off lists
The production of material take-off lists can also be augmented by AI tools today.
For example, AI-tools exist that are able to produce material take-off lists in a matter of minutes by analysing construction drawings, detecting and labelling architectural and structural elements, measuring dimensions and calculating associated quantities. Required material type, quantities, locations and specifications can then be identified and organised into material take-off lists in a fraction of the time it would take a human engineer. Human oversight, however, is likely to remain an important factor in this process until these tools are tried, tested and proven.
Existing design tools are also being integrated with AI. For example, BIM has long since been used in the construction industry to assist with design development. BIM is being integrated with AI which may, for example, assist an architect by creating and testing hundreds or thousands of alternative designs in response to the architect changing the design parameters on a live basis. This can assist with ensuring a design meets the regulatory framework of the jurisdiction in which a project is being carried out, for example fire safety regulations under or in connection with the Building Safety Act 2022 in England.
Augmented-reality headsets can also be used by designers to help visualise concepts by overlaying 3D models with a physical environment. This can assist not only with basic design but with site analysis, including understanding how a 3D model may interact with ground conditions, topography or drainage patterns.
AI-enhanced quality control and site monitoring
Traditionally, quality control has relied largely (if not solely) on human inspectors and site supervisors to visually assess workmanship and materials. That process can be very time consuming, expensive, prone to human error and generates a huge amount of documentation that can be unwieldy.
The construction industry appears to be implementing an increasing number of AI-driven methods to streamline this process through the use of cameras, drones and/or sensors that are able to capture and analyse images and data on a live basis during a project's construction phase. AI tools exist that can analyse this documentation and instantly detect defects, structural issues or general deviations from a project's design.
AI in scheduling and project management
AI can be used to improve and augment traditional project scheduling and risk management. By integrating AI into existing scheduling programmes, (such as Primavera P6) AI is able to analyse past project schedules and assist project planners to optimise current project schedules by identifying any pinch-points and suggesting solutions to reduce delays.
By learning from past projects, AI can assist with the identification of potential critical delays it deems likely and suggest ways a schedule may be optimised as a result, including an assessment of proposed manpower.
AI may also be used to assist with the early resolution of delay or disruption disputes, saving time and cost.
Disruption events are notoriously difficult to prove, however, AI may act as an objective evaluator of the effect of disruptive events based on live performance data, schedules, correspondence and other available data. Project managers may use that analysis to make a more informed decision on contractor claims.
AI for health and safety compliance
Given the importance placed on health and safety in construction projects around the world, it seems inevitable that construction companies globally are likely to adopt AI to improve health and safety on site.
AI can be implemented to ensure greater compliance with on-site safety practices. For example, fixed and mobile cameras, integrated with image recognition AI technology can be used to monitor worksites in real time, identifying issues such as workers without PPE or dangerous conditions such as trip hazards or areas where safety netting or guard rails have not been installed. Wearable technologies such as smart helmets and smart vests can also be used to monitor worker behaviours and environmental conditions. Supervisors can be informed of any discrepancies in seconds, rather than having to rely on human oversight.
AI driven robotics and automation
AI-driven training programmes can assist with skill development by personalising the training experience and using augmented reality to provide immersive, simulated environments for training purposes.
AI-powered robotics can be used to automate labour-intensive construction tasks. For example, using autonomous robots to directly mark architectural and engineering designs onto floors with speed and accuracy. Similarly, worker-controlled robots are available that use machine vision to complete drywall tasks faster than more traditional means.
AI in dispute resolution: a glimpse ahead
As the use of AI increases in construction projects, we may start to see the emergence of new types of disputes, such as those discussed below:
IP ownership
Where AI is used to assist with the production of construction designs, it may be that an AI system produces a new, innovative type of structural design that holds inherent value. Disputes may arise over who owns the IP to that innovative design – the contractor utilising the AI? The AI proprietor? The employer? Clear contractual terms between parties should be adopted to account for this possibility.
Care will need to be taken with the data used to train AI; and generated results considered with an understanding of the risks of the AI having copied or otherwise re-used IP owned by a third party.
Liability for error
What happens when an AI system provides inaccurate data, "hallucinates" or makes faulty recommendations leading to, for example, project delay or a safety incident? The obvious answer may be that the party utilising the AI should be responsible, but what if the error was made because of incorrect (or allegedly incorrect) or inadequate data provided by another party? Complex disputes over responsibility for such error may arise, requiring detailed factual and expert analysis.
Another factor to be considered is whether the proprietor of the AI system holds any responsibility, such that any claim(s) may be passed on.
Employment
Certain jurisdictions have national legislation that requires a specific proportion of a project's workforce to be hired from the jurisdiction where the project is located. For example, Saudi Arabia has a "Saudization" scheme that requires companies operating in Saudi Arabia to hire a specific percentage of Saudi nationals to the project's workforce.
In construction projects in the future, an increasing number of site staff may be replaced by AI – for example AI-driven robots can lay bricks or control concrete mixers and pourers with accuracy and precision with no requirement for rest. Coupled with AI-powered quality or safety inspection, that may result in a significant reduction of the number of on-site staff required. Where a contractor's project team is headquartered overseas, this type of AI may impact its willingness to meet such employment quotas.
Similarly, in extreme circumstances AI may lead to mass workforce redundancies, potentially leading to claims for unfair dismissal.
Confidentiality, data or privacy breaches
Construction companies that train AI systems on past projects for past clients may run the risk that an AI system utilises confidential or patented design technology (accessed through its training) when producing a design on a future project. It may not necessarily be obvious to the human designer operating the AI that the design produced includes such technology and may inadvertently implement it on a future project without the required licences.
More broadly, there are consistent concerns among the legal profession about data or privacy breaches from the use of AI. For example, in 2017 the Royal Free NHS Trust was found to have unlawfully shared patient data with Google DeepMind where Google DeepMind had used AI in an effort to help detect signs of early kidney disease by processing data from past patients.
Preparing for an AI driven future
AI is no longer a distant concept, and it is worth noting that it will play an increasingly significant role in dispute resolution, and across the legal sectors.
Ultimately, in a world of hourly rates, clients expect their lawyers to use technology to enhance efficiency and reduce time spent, and therefore, cost, without compromising on quality. From document management and multilingual translation to workflow optimisation, AI offers tangible benefits to the dispute resolution process.
AI is here to stay, and its influence is set to grow exponentially. The vast majority of academic and scientific debate surrounding the influence of AI is centred around the question of "when", not "if". The international construction industry can (and, no doubt, will) take huge benefit from its use – not least from a health and safety standpoint – but must also be aware of the risks involved.
Construction companies and in-house legal teams must proactively engage with AI's capabilities and risks. This includes implementing robust governance frameworks, ensuring clear contractual terms, and fostering collaboration between legal, project, and technical teams. As disputes increasingly intersect with AI, whether through its use or misuse, strong document management and informed workflows will be critical.
Ultimately, embracing AI with strategic foresight will position organisations to not only resolve disputes more effectively, but also to prevent them. The question is no longer if AI will transform the industry, but how prepared we are to harness its potential. For those looking for guidance on how best to manage and engage with AI's capabilities and risks in construction projects, the Construction Team at Gowling WLG is well placed to help.
Read the original article on GowlingWLG.com
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
