Artificial intelligence is becoming an increasingly important part of digital transformation across the architecture, engineering, construction and operations (AECO) sector. Developers, consultants, contractors, technology vendors and specialist service providers are experimenting with AI to improve areas such as design, project planning, project controls, construction delivery and asset operations.
However, while AI experimentation is growing, widespread adoption across the construction industry remains limited. According to the RICS Artificial Intelligence in Construction Report 2025, based on a global survey of more than 2,200 respondents, 45% of organisations had not implemented AI, while 34% were still in the early stages of pilot projects. Less than 1% had achieved fully embedded, organisation-wide AI adoption.
These figures highlight a broader challenge for the industry: moving beyond individual AI experiments and finding ways to scale technology while creating measurable value at the project level.
Moving From Company Adoption to Project Value
Most organisations currently explore AI within their own areas of responsibility. Consultants may develop AI applications for design and engineering, contractors may focus on construction and field operations, and developers may investigate opportunities throughout development, delivery and asset operations.
These individual initiatives can create value for individual organisations, but they do not necessarily result in an integrated AI ecosystem across an entire construction project.
A typical construction project involves numerous organisations, each bringing its own processes, information, expertise and technology. If AI adoption remains limited to individual organisational boundaries, opportunities to address shared project challenges may be missed.
This does not mean that clients should control how consultants, contractors or other partners develop their internal AI capabilities. Those decisions remain the responsibility of each organisation.
At the project level, however, the client occupies a different position. The client typically owns the investment, establishes project objectives and maintains a long-term interest in the project’s outcome. This gives the client an important role in establishing AI governance across the supply chain.
Rather than centralising every AI application, the client can act as an orchestrator, helping different stakeholders coordinate their AI initiatives around common project objectives.
Start With the Use Case, Not the Technology
One of the common challenges with emerging technology is beginning with the technology itself and then searching for ways to apply it. A more practical approach is to begin with an actual business or project challenge and determine whether AI can provide a credible solution.
Research from Autodesk in 2025 illustrates this challenge. The research found that 47% of construction leaders considered identifying the right AI use cases to be a major or moderate concern, while only 32% said they were approaching or had achieved their AI goals.
Potential AI applications should therefore be evaluated using practical criteria, including:
- Expected business impact and measurable benefits
- Implementation complexity and scalability
- Availability, quality and readiness of data
- Risk, governance and human oversight requirements
The next step is to turn promising use cases into clearly defined business cases. Each business case should establish a baseline describing the current situation, identify expected benefits and define measurable success criteria.
This changes the fundamental question from “Where can we use AI?” to “Where can AI create measurable project value?”
Bringing the Construction Supply Chain Together
Client leadership does not eliminate the importance of contractors, consultants, technology companies or other supply chain participants. Their involvement is essential to successful AI implementation.
Design-related applications may depend on consultants’ technical expertise and project information. Construction applications may require contractors’ workflows, site knowledge and field data. Technology and solution providers can contribute platforms, integration capabilities and implementation expertise.
The opportunity is to bring these capabilities together around clearly selected project-level use cases rather than allowing multiple disconnected AI experiments to develop independently.
This leads to an important principle: AI ownership should follow accountability.
For each AI use case, stakeholders should clearly define who is responsible for the business outcome, technology, data, implementation, validation, human oversight and adoption.
A suitable governance structure should also address areas such as data security and AI data sovereignty, intellectual property rights, applicable regulations and the boundaries of automated decision-making.
From AI Deployment to Measurable Value
Launching an AI pilot does not automatically mean that a project has achieved successful transformation.
If an AI business case is intended to reduce review times, improve forecasting accuracy, minimise rework, increase quality or improve productivity, the results need to be measured against the original baseline.
The client can establish the overall framework for measuring value, while supply chain stakeholders contribute to implementation, measurement and continuous improvement.
A practical process can include four stages:
- Assess the use case and define the business case
- Establish ownership and deploy collaboratively
- Measure realised value against the baseline
- Scale successful applications while adjusting or discontinuing those that do not deliver the expected results
This approach helps close the gap between simply deploying AI and demonstrating that the technology has generated tangible project value.
What the Next Stage of AI Maturity Could Look Like
AI experimentation is expanding across the AECO sector, but the industry’s next challenge is turning experimentation into consistent and measurable outcomes.
Greater AI maturity is not necessarily about every organisation adopting more AI independently. Instead, it can involve clients establishing appropriate governance at the project level and enabling consultants, contractors, technology vendors and solution providers to collaborate around shared objectives.
AI capabilities may remain distributed across the construction supply chain, but project-level accountability for outcomes remains important. By coordinating stakeholders, defining responsibilities and measuring results, clients can help ensure that AI adoption contributes to meaningful project performance.
Conclusion
AI has significant potential across construction, from design and planning to project controls, field operations and asset management. However, technology alone does not guarantee better project outcomes.
The focus needs to shift from isolated AI pilots toward clearly defined use cases, measurable business cases, coordinated implementation and effective governance. Clients are positioned to play an important coordinating role because they establish project objectives and remain accountable for the overall investment and outcome.
By leading governance without unnecessarily centralising technology decisions, clients can create an environment where different members of the construction supply chain contribute their expertise while working toward shared, measurable results.
FAQs
Clients typically establish project objectives and oversee the overall investment, giving them a strong position to coordinate AI governance at the project level. Their role can focus on orchestration rather than controlling how individual organisations use AI internally.
AI use cases should be evaluated based on factors such as potential business impact, measurable benefits, implementation complexity, scalability, data availability and quality, risk, governance and human oversight requirements.
Projects can establish a baseline before implementing an AI solution and then measure outcomes against defined success criteria. Metrics may include reduced review time, improved forecast accuracy, lower rework, increased quality or improved productivity.