Dave Philp FCIOB, chief value officer, Bentley Systems and chair of the CIOB Innovation & Digital Advisory Panel, explores how AI in construction has progressed over the years
Artificial intelligence (AI) has moved remarkably quickly from boardroom curiosity to practical reality. Just a few years ago, discussions about AI in construction focused on futuristic concepts, autonomous machines, and speculative promises. Today, AI is already embedded within many of the processes we use every day, helping us design, plan, deliver, and operate built assets more effectively.
One message the Chartered Institute of Building (CIOB) AI Playbook consistently emphasised is that AI should not be viewed as a replacement for construction professionals.
Instead, we should think of it as a new colleague, a digital assistant capable of handling repetitive tasks, analysing vast datasets, identifying patterns, and supporting better decision-making. The opportunity is not simply to do the same things faster. It is to fundamentally improve how we deliver the built environment.
Construction remains one of the world’s largest industries, yet productivity growth has historically lagged other sectors. Projects are increasingly complex, margins remain tight, skilled labour shortages persist, and sustainability requirements continue to intensify. AI arrives at a time when the industry needs new approaches.
The question is no longer whether AI will impact construction. The question is how quickly organisations can harness it responsibly and effectively.
Moving beyond the chatbot phase
Much of the public discussion around AI has centred on generative AI tools that can create text, images, and content. While these capabilities are often impressive, they represent only the visible tip of a much larger opportunity. The real value emerges when AI becomes integrated into end-to-end construction workflows.
Rather than a prompt simply helping someone write a report faster, AI can support an entire process. It can analyse project information, identify risks, generate recommendations, track progress, highlight deviations, and provide real-time insights.
Recent industry research suggests a significant proportion of non-physical work activities in architecture, engineering, and construction contain tasks that could be augmented or partially automated. The most successful organisations will not be those that deploy isolated AI tools. They will be the ones who redesign workflows around human-AI collaboration.
In many respects, we are witnessing a transition from AI as a productivity tool to AI as a project teammate.
Design optimisation at scale
One of the clearest examples of AI delivering value is in design.
Traditionally, engineers and designers evaluate a limited number of options before selecting a preferred solution. AI-powered generative design systems can explore thousands of permutations within defined parameters, helping teams identify options that balance the trade-offs of cost, carbon, constructability, performance, and operational efficiency.
This does not remove human judgement – it expands the solution space available to designers, enabling professionals to make more informed decisions.
Infrastructure owners are increasingly using AI-enhanced modelling and simulation tools to optimise layouts, assess environmental impacts, and improve asset performance before construction begins. This ability to test scenarios digitally reduces uncertainty and enables better decisions earlier, when changes are less expensive and more impactful.
Predicting risk before it happens
Risk management has always been central to successful project delivery.
Construction projects generate enormous amounts of information through schedules, parametric models, RFIs, site reports, sensor feeds, cost data, and progress updates. Historically, much of this information remained fragmented across multiple systems.
AI excels at finding patterns within large datasets that humans may overlook.
Project teams increasingly use machine learning models to identify emerging schedule risks, predict potential cost overruns, and highlight areas where projects drift from baseline plans. Rather than discovering problems after they occur, teams can intervene earlier.
This predictive capability is particularly important given the increasing scale and complexity of modern infrastructure programmes. As projects become more data-rich, AI becomes more valuable as an early warning system.
Improving safety outcomes
Construction remains a high-risk environment, making safety and well-being one of the most promising areas for AI adoption.
Computer vision systems can analyse site imagery and video feeds to identify unsafe conditions, monitor compliance with safety procedures, and flag potential hazards.
Wearable technologies combined with AI analytics can help organisations better understand workforce movements, fatigue indicators, and environmental conditions.
Most importantly, AI enables organisations to shift from reactive safety management to predictive safety management. By analysing historical incidents, near misses, and operational data, AI can identify leading indicators that help prevent accidents.
Every injury avoided represents a meaningful return on investment.
Digital twins become intelligent
One of the most exciting developments is the convergence of AI and digital twins.
Digital twins provide dynamic virtual representations of physical assets. When combined with AI, they become significantly more powerful.
AI can analyse operational data flowing through a digital twin and identify inefficiencies, predict maintenance requirements, optimise performance, and simulate future scenarios.
Consider a water network, rail system, or data centre. AI-enhanced digital twins can continuously monitor performance, detect anomalies, predict failures, and recommend interventions before disruptions occur.
The rise of Agentic AI
A major emerging trend is the move toward agentic AI. Most current AI systems respond to prompts. Agentic AI goes further by pursuing objectives, coordinating tasks, and orchestrating workflows.
Imagine a project controls assistant that can review programme data, identify delays, draft mitigation actions, notify stakeholders, and update dashboards automatically. Or a design coordination assistant capable of detecting clashes, proposing solutions, and initiating review workflows.
Again, these systems do not replace project teams. They augment them.
Construction remains fundamentally a people business requiring critical thinking and judgement, leadership, collaboration, negotiation, and accountability. However, agentic AI can remove substantial administrative burden, allowing professionals to focus on higher-value activities.
This shift may prove as significant as the industry’s original transition from paper-based processes to digital workflows.
Evidence from early adoption
Across the sector, CIOB has seen positive outcomes that continue to grow.
Contractors and their supply chain are using AI-powered planning tools to improve schedule reliability and resource allocation.
Engineering firms are deploying generative design systems to accelerate option development and improve project outcomes.
Owners and operators are leveraging AI-enabled digital twins to optimise asset performance and reduce operational costs.
Many organisations report substantial reductions in time spent on repetitive administrative activities such as document review, information retrieval, reporting, and compliance checking.
Perhaps the most significant benefit is not any single efficiency gain. It is the cumulative impact of thousands of small improvements occurring across the project lifecycle.
When design teams make better decisions, planners identify risks earlier, site teams improve productivity, and operators optimise performance, the combined value becomes transformational.
Trust, governance and human oversight
While enthusiasm is justified, responsible adoption remains essential.
The CIOB AI Playbook emphasises four critical principles: trust, transparency, explainability, and governance. These principles are becoming increasingly important as AI systems become more sophisticated.
Construction professionals remain accountable for project decisions. AI can provide recommendations, but responsibility cannot be delegated to an algorithm.
Organisations must establish clear governance frameworks, ensure data quality, validate outputs, and maintain human oversight. The goal is not blind automation. It is trusted augmentation.
Those who balance innovation with responsibility will realise the greatest long-term benefits.
The future is collaborative
AI will not build projects on its own. It will not replace engineers, project managers, architects, surveyors, or constructors. Instead, it will amplify human capability.
The organisations that succeed will treat AI not as a technology initiative but as a business transformation initiative. They will focus on workflows rather than tools, outcomes rather than experimentation, and people rather than algorithms.
Construction has always been an industry that transforms ideas into reality. AI provides an opportunity to do that faster, safer, more sustainably, and with greater confidence.
The future of construction is not human versus machine. It is human and machine working together to deliver better outcomes for society.
The post AI in construction: From productivity tool to project teammate appeared first on Planning, Building & Construction Today.
