Karl Smith at Creative ITC explains how architectural practices can move from experimenting with AI to strategic integration, by leveraging structured data for measurable outcomes – thereby building their capability to serve future needs.
More than half of architecture professionals are already using AI in their operations, with many more planning adoption in the coming years. Practices are seizing the powerful opportunity to accelerate creativity, support faster design exploration and improve project outcomes. These benefits are increasingly critical as the industry faces mounting pressure to design and build smarter to meet the rising demands of a growing global population projected to reach 10 billion by 2060.
However, artificial intelligence is moving faster than the architecture industry’s ability to restructure itself around it. Many practices are struggling to realise their full potential. Legacy IT systems, siloed data and AI capabilities make it difficult to deploy AI effectively, exposing firms to unreliable outputs, delayed adoption and higher operational risk. To move beyond experimentation, firms need a clear understanding of AI’s real world value and the barriers that limit its impact.
From curiosity to commercial execution
AI is already reshaping architectural workflows. According to the NBS 2025 Digital Construction Report, adoption continues to accelerate across design and delivery teams. 42% of architecture professionals are using AI tools today, with a further 37% planning adoption within five years. Initial fears that AI would replace creativity have largely subsided. Instead, AI is increasingly augmenting it by expanding design exploration, accelerating feasibility studies and supporting faster decision making.
However, the real transformation has now moved beyond visualisation into operational intelligence. Across the asset lifecycle, AI is being integrated with IoT and digital twins to create closed-loop decision systems. IoT gathers real time telemetry. AI translates that data into predictions and risk signals. The digital twin provides contextual simulation to test outcomes before action is taken. Insights generated during delivery and operation then cycle back to inform future design decisions.
Competitive advantage is being built in data-backed execution. Indexed BIM and IFC datasets are enabling delivery teams to ask the model direct natural-language questions. Computer vision compares site imagery, LiDAR and scans against federated models to detect variance early. Historical programme data is training schedule risk models that forecast delays before they surface commercially. Pre-design costing models are becoming more accurate through structured portfolio data.
When grounded in structured, high-quality data, AI stops being speculative and delivers real world operational impact. This is what moves margins.
Hidden risks: why AI pilots stall
While most practices are experimenting with AI tools, fewer than 1% of architecture firms report organisation-wide AI integration. A key reason for this gap is that many practices approach AI as a standalone innovation initiative rather than as an organisation wide enabler. This means many firms neglect the robust data groundwork on which AI success relies.
AI does not fix weak digital foundations – it exposes them. AI built on fragmented foundations amplifies confusion. Siloed legacy systems result in AI exposing interoperability weaknesses. If governance is unclear, AI introduces risk rather than efficiency.
Legacy IT remains a significant barrier. Outdated, batch-oriented IT environments struggle to support real-time AI workflows. Interoperability gaps between design platforms, cost systems and project controls limit insight. Cloud migrations undertaken
without architectural planning often introduce further complexity through unexpected egress costs, performance constraints and sovereignty concerns.
At the same time, firms also face growing risks around security and intellectual property. Feeding proprietary design data into external AI models without clear governance can compromise competitive advantage. Large models may unintentionally replicate protected content. These risks are leading many firms to move away from off-the-shelf tools to develop more secure proprietary systems based on their firm’s own data and platforms. Without defined guardrails, firms risk commercial exposure in pursuit of efficiency.
AI integration demands operational redesign
True AI maturity begins with a simple but often avoided question – what commercial outcome are we trying to improve? Architecture firms must first define the results that matter most – increased bid success rates, reduced design-stage rework, improved delivery margins, accelerated programme certainty, reduced insurance exposure, improved utilisation and fee recovery.
AI that is not anchored to top-line growth or bottom-line protection will remain experimental. Industry specialist partners can help practices cut through AI hype and focus on commercial impact. With that direction set, firms can build a pragmatic roadmap that prioritises the right use cases, data sets and tools. When aligned to measurable KPIs, AI investment becomes strategic rather than tactical.
To support scalable AI, firms must modernise their digital foundations. Workloads must be placed in the right environments, whether on-premise, private or sovereign cloud, public cloud or edge. A cloud-smart strategy balances governance with agility, enabling secure data mobility while maintaining compliance
and performance.
Data must be central, structured and governed. Design archives, schedules, project decisions and cost histories should not sit in isolated silos. When AI models are anchored in structured, portfolio level knowledge, outputs become reviewable, traceable and commercially defensible. This is the difference between speculative prompting and institutional intelligence.
However, even the best laid IT plans can go awry when firms underestimate the human element. Even when practices define the right problem and select the right tools, adoption often stalls at the point of operational integration. AI only creates value when teams know how to use it effectively.
Effective adoption requires training teams on how prompts influence outcomes, teaching users how to structure inputs for accuracy, establishing clear usage protocols, embedding AI into existing workflows, and creating feedback loops to refine output quality. AI capability building must sit alongside technical deployment. Firms that treat AI education with the same seriousness as BIM training will move faster and safer than those relying on informal experimentation.
Harnessing AI for long-term advantage
AI integration is ultimately a leadership test – and one that will expose inefficient operating models across the architecture industry. The firms that will lead the next decade of digital construction will move on from experimenting with tools and focus instead on building the cleanest data environments, defining the clearest commercial objectives and investing in capability development. AI transformation is not about chasing innovation headlines. It is about building operational intelligence that compounds over time.
For smaller practices, this shift represents a significant opportunity. With strong foundations and disciplined governance, AI can level the playing field, allowing lean teams to compete at scale without matching headcount. For larger firms, the stakes are higher. Marginal gains in delivery efficiency translate into substantial commercial advantage across global portfolios.
AI holds enormous potential across the architecture sector, but firms will only realise its full value if they deploy it with intention and discipline. Firms must define the commercial problem first, align AI initiatives to measurable outcomes, strengthen digital and data foundations, establish governance before scale, invest in structured user capability and training, and embed AI into operational workflows rather than treating it as an overlay.
Architecture firms that move from experimentation to disciplined execution will be the first to reap rewards far greater than efficiency. Those that act now will be best positioned to turn AI into a true strategic advantage, reshaping how projects are conceived, delivered and optimised across the entire asset lifecycle and delivering the infrastructure the future demands.
Karl Smith is director at Creative ITC