AI agents are rapidly entering enterprise operations, yet many organisations remain unaware of their hidden costs.
Across the digital construction sector, AI agents are shifting from isolated trials to core operational infrastructure. In 2026, enterprises are moving beyond simple chatbots and deploying autonomous agents to handle complex work. Hiring teams are testing screening agents, legal teams are using agents to review contracts, and operations teams are connecting agents to internal tools and project data.
In construction and the built environment, this trend is accelerating. AI agents are being applied to procurement workflows, invoice processing, design coordination, compliance checks, and asset data management. However, governance, identity management, and access control frameworks have not evolved at the same pace. As a result, many organisations lack clear visibility into where AI agents run, what systems they access, and who is accountable for them.
This is not a future risk. The World Economic Forum reports that 82% of organisations plan to integrate AI agents within the next one to three years, signalling that agent-based workflows are moving rapidly beyond pilot projects into enterprise scale.
The growing AI governance gap in enterprise environments
The absence of structured AI governance is becoming a material risk. An IBM study found that 63% of organisations that experienced security breaches either lacked AI governance policies or were still developing them. Even where policies exist, fewer than half include strict approval processes. Only a small proportion actively audit for unauthorised AI use, and 61% do not use dedicated AI governance tools.
Recent CISO research reinforces this picture. In a survey of 235 senior security leaders, 92% reported lacking full visibility into AI identities, including non-human accounts used by agents, copilots, and assistants. Additionally, 75% confirmed that unsanctioned AI tools were already operating within their environments.
For digital construction organisations managing sensitive commercial data, intellectual property, and operational systems, this lack of oversight introduces both financial and security exposure.
Overlapping AI agents and the rise of hidden operational costs
When teams deploy AI agents independently, duplication becomes common. Finance teams may build an agent to process invoices, while procurement teams create a separate agent to achieve the same outcome. Marketing, sales, and project delivery teams may each deploy their own customer or asset data agents.
Without a central inventory or shared visibility, these overlaps remain undetected. The result is duplicated workflows, fragmented tooling, and escalating costs associated with API usage, compute consumption, and licensing.
Žilvinas Girėnas, head of product at nexos.ai, an all-in-one AI platform for enterprises, explains the challenge: “AI agents are becoming part of day-to-day operations. Leaders aren’t worried about the agents themselves, but about losing visibility once they invoke external functions, APIs, or services, actions that take, query databases, or shift data across systems. The question shifts from ‘Which AI tool are employees using?’ to ‘Which agents exist, what identity do they run under, and what can they access?’ With an inventory, clear ownership, and audit logs in place, companies can scale agents with control and confidence.”
He adds, “When several teams build their own agent for the same job, you do not get more value. You get duplicated workflows and fragmented spend. Seats, runs, and queries show up across different tools, and your API costs rise.”
AI cost pressure and margin erosion
Research shows that 84% of companies believe AI costs are negatively impacting gross margins by more than six percentage points. For SaaS and digital platform businesses operating at scale, this represents a substantial financial impact. Construction technology providers and digitally mature contractors face similar pressures as AI adoption increases.
The difficulty lies in attribution. Teams struggle to identify which AI agents are consuming resources, which are delivering value, and which are quietly inflating costs. Without visibility into runtime behaviour and usage patterns, financial forecasting becomes unreliable.
“The paradox is that the same tools that help teams work faster also risk breaking cost control,” says Girėnas. “Speed and visibility must go together, or it leads to chaos and unexpected expenses.”
Common governance failures organisations encounter
Governance breakdowns typically appear in four ways. First, organisations lack a clear understanding of which agents are active and who approved them. Second, overlapping workflows lead to wasted spending and duplicated effort. Third, agents operate across systems without sufficient security checks. Finally, finance teams cannot forecast or manage costs due to limited insight into actual AI resource consumption.
Over time, these failures accumulate, turning operational efficiency gains into recurring financial surprises.
“The goal isn’t to slow down your fast-moving teams,” Girėnas explains. “It’s to let them build quickly while keeping track of what they’re building, who they’re giving access to, and what it costs. This is the difference between a scalable platform and infrastructure that might become a liability.”
What digital construction leaders should do now
Agent sprawl escalates quickly once organisations scale. Without visibility, costs compound quarter after quarter. Girėnas recommends four practical guardrails.
First, create a comprehensive AI agent inventory. A central registry documenting every agent, owner, and access level establishes baseline visibility and highlights duplicated workflows early.
Second, separate build-time freedom from runtime control. Teams should be free to design and iterate quickly, while a shared control plane manages identity, permissions, and logging.
Third, treat AI agents as employees rather than scripts. Define roles, limit privileges, and require sandbox testing before granting access to sensitive systems.
Finally, connect agents directly to cost signals. Linking usage and spend to individual agents enables early intervention and informed decision-making.
Without visibility, AI agent sprawl quietly erodes margins, security, and trust, making governance, cost tracking, and accountability essential foundations for scalable AI adoption in digital construction environments.
Read next: Building marketing value through Top 25 Tech recognition – Build in Digital
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
