Digital construction and smart cities are at the centre of many GCC countries plans, yet many projects fail to move beyond pilots. Ibrahim Imam, CEO and Co-founder of PlanRadar, explores how fragmented documentation, inconsistent workflows and unstructured project data remain one of the biggest hidden barriers to scalable AI adoption in construction.
Across the GCC countries, Artificial Intelligence is moving quickly from theory to boardroom priority. The UAE’s National Strategy for Artificial Intelligence 2031 positions AI as a core driver of productivity across infrastructure and the built environment, while Saudi Arabia’s Vision 2030 places digital construction and smart cities at the centre of its US$1 trillion+ development pipeline. Developers, contractors and asset owners are increasingly exploring how AI could support faster delivery, improved safety and better decision-making across increasingly complex construction portfolios.
Yet despite the momentum, many projects are struggling to move beyond pilots and isolated use cases.
Today, construction projects generate vast volumes of digital information. Mobile inspections, progress photos, issue logs, daily reports and handover documentation are widely used on jobsites. However, on many projects, the information is still fragmented across disconnected systems, personal workflows and multiple versions of the truth.
This fragmentation sits at the centre of the AI conversation in construction. While teams ask what AI can do for their projects, a quieter and more important question remains: is the information reliable enough for AI to learn from and act on?
On live projects, three issues continue to surface. First, information is often scattered. A checklist may sit in one system, photos in another, approvals in email threads and drawings in shared drives. Second, documentation standards vary by individual, trade or shift, making it difficult to compare or trust records. Third, data is growing faster than teams can organise it, leading to more time spent searching for information, resolving disputes or reworking tasks that should already be complete.
These challenges are particularly visible in large-scale GCC projects, where multiple contractors, consultants and subcontractors operate simultaneously across phases and packages. In such environments, even small inconsistencies in documentation multiply quickly. Studies indicate that 27% of AI projects in Saudi Arabian businesses are stalled or cancelled due to unorganised or untrusted data. When there are competing versions of site reality, AI systems cannot reconcile them. They will analyse whatever data they are given, including gaps, duplicates or outdated records. The result is insight without confidence.
Projects that perform well tend to agree early on how issues are logged, how locations are defined, how evidence is captured and what ‘closed’ means. These decisions sound simple, yet they are critical on fast-moving sites where processes must work on busy days, not only when schedules are calm.
AI begins to add real value when this foundation is in place. Practical use cases already exist across the region, particularly where information is repetitive and time sensitive. On a large mixed-use development, for example, consistent defect logging allows AI to group recurring issues by location or trade, helping site managers prioritise areas of risk. Structured photo documentation enables patterns to be identified across floors or zones, highlighting quality or safety concerns earlier than manual reviews would allow. Clear ownership and close-out standards make it easier for systems to flag overdue items or missing evidence before they escalate into disputes or delays.
This is why the most AI-ready projects in the GCC today are not those adopting the most tools, but those applying digital workflows consistently across their sites. High-performing teams typically rely on a single agreed process for logging and closing issues, one shared set of current documents and clear naming and tagging rules that make information easy to find later. These choices reduce friction day-to-day while creating a reliable record that can be analysed over time.
Digital platforms such as PlanRadar support this approach by helping teams capture issues, inspections and evidence in a consistent, mobile-first way. When information is recorded at the source and stored in a single system, teams spend less time reconciling data and more time resolving work. Just as importantly, they create a structured dataset that can support future AI applications with confidence.
Looking ahead, AI’s role in construction will become more operational and less experimental. Over the coming years, we can expect greater use of AI to support predictive quality management, automated risk identification and smarter handover processes aligned with regulatory and ESG requirements in the UAE and Saudi Arabia. However, these capabilities will only deliver value where projects have invested in disciplined documentation and repeatable site workflows.
