Artificial Intelligence (AI) has rapidly moved from experimentation to strategic priority. Organizations across government, infrastructure, construction, transport, energy and financial services are investing heavily in AI technologies to improve decision making, automate processes and unlock new efficiencies. Yet despite growing investment, many organizations are struggling to move beyond isolated pilots and demonstrations into scalable enterprise AI. The problem is rarely the AI technology itself.
In most cases, the challenge lies in the quality, consistency and governance of the underlying organizational data.
As Deloitte notes, only
“42% of companies believe their strategy is highly prepared for AI adoption”
meaning many organizations continue to face challenges scaling AI across the enterprise because foundational capabilities remain immature.
Additionally, Deloitte state,
“legacy data and infrastructure architectures cannot power real-time, autonomous AI. As AI capabilities extend beyond software into devices, machinery, and edge locations, organizations need to evaluate if their technology foundations are ready to support potential physical AI deployments”.
Similarly, KPMG highlights that
“poor data governance remains one of the primary barriers inhibiting AI initiatives”.
This is becoming one of the defining issues of enterprise AI adoption. AI can only be as effective as the information environment it operates within.
The AI ambition gap
Many organizations today possess large volumes of data. However, very few possess data environments that are genuinely AI ready. Over time, information has become fragmented across disconnected systems, spreadsheets, email trails, local databases and departmental applications. Critical project information often exists in different formats, follows different standards and lacks clear ownership.
As organizations attempt to introduce AI capabilities into this landscape, problems quickly emerge. Teams encounter inconsistent reporting structures, duplicate records, conflicting project data, incomplete program histories and limited visibility across portfolios. AI tools may generate insights, but the underlying information is often unreliable or impossible to interpret consistently. This creates what many organizations are now experiencing as the AI ambition gap. The organization has strategic ambitions for AI but lacks the operational and data foundations required to support it at scale.
A recent KPMG report noted that
"traditional governance approaches are increasingly inadequate in an AI driven environment because they were designed for structured, human led data consumption rather than dynamic, enterprise-wide AI usage."
In practical terms, this means organizations often attempt to deploy advanced AI capabilities onto fragmented operational environments that were never designed to support them.
Why foundational data matters
There is growing recognition that AI readiness is fundamentally a data challenge before it becomes a technology challenge. Without these foundations, AI outputs become difficult to trust, difficult to scale and difficult to govern. This issue becomes particularly important in project intensive organizations where decisions rely heavily on complex delivery data.
Large programs generate enormous volumes of operational information across cost, schedule, risk, resource management, change control, procurement and governance activities. If this information is inconsistent or fragmented, AI systems struggle to produce reliable insights.
For example, an AI model attempting to forecast delivery risk across a capital portfolio may encounter:
Different naming conventions between business units
Inconsistent risk scoring methodologies
Missing project baseline data
Unstructured status reporting
Incomplete program histories
Multiple disconnected reporting tools
The result is not intelligent automation. The result is amplified inconsistency.
AI does not eliminate weak data foundations. In many cases, it exposes them faster and at greater scale.
Gartner recently cited,
through 2026, organizations will abandon 60% of AI projects unsupported by AI-ready data.
Organizations achieving the greatest AI success are typically those that completed the hard operational and governance work first.
The growing importance of governance
As AI adoption accelerates, governance is rapidly becoming a board level concern. This extends beyond regulatory compliance. Governance increasingly determines whether AI outputs can be trusted, audited and operationalized safely across the organization.
KPMG recently stated that
"organizations must unify AI and data governance under a shared governance framework in order to create transparency, enforce standards and reduce risk."
This is particularly relevant within government and regulated industries where accountability, transparency and traceability remain critical. This challenge is becoming increasingly visible as organizations move from experimentation into operational AI deployment.
Deloitte research suggests
"many organizations are accelerating investment in AI agents and autonomous systems while governance maturity continues to lag behind adoption rates. Only one in five (21%) companies surveyed report currently having a mature model for governance of autonomous agents."
The implication is clear. AI maturity is no longer simply about deploying models or tools. It is about creating governed operational environments capable of supporting intelligent decision making at enterprise scale.
Why project environments are especially vulnerable
Project intensive organizations often face additional complexity because delivery information is inherently dynamic. In many organizations, project information remains heavily manual and dispersed across disconnected reporting structures.
This creates several challenges for AI adoption:
Lack of standardization
Different projects frequently use different reporting formats, governance structures and delivery methodologies. This makes portfolio level analysis difficult and limits AI’s ability to identify meaningful patterns.
Poor data lineage
Many organizations struggle to trace how information has been created, updated or approved across the project lifecycle. Without clear lineage, AI generated recommendations become difficult to validate.
Limited cross portfolio visibility
AI systems are most effective when they can analyze large, connected operational datasets. Siloed project environments restrict the ability to identify trends, dependencies and emerging delivery risks across portfolios.
Inconsistent governance
If governance standards differ between programs, AI systems inherit those inconsistencies. This reduces trust in automated reporting and forecasting outputs.
Unstructured operational knowledge
Large volumes of project intelligence remain trapped inside emails, presentations, spreadsheets and narrative documents. Without structured operational frameworks, much of this information remains inaccessible to enterprise AI.
The role of PPM in AI readiness
This is where project and portfolio management (PPM) begins to play an increasingly critical strategic role. Historically, PPM has often been viewed primarily as a delivery management discipline focused on governance, reporting and oversight. Modern PPM environments create the structured operational frameworks that AI systems depend upon.
They establish:
Common delivery standards
Consistent project structures
Integrated governance processes
Portfolio wide visibility
Controlled reporting environments
Reliable historical delivery records
This creates a significantly stronger platform for enterprise AI adoption. Importantly, this is not about using AI simply to automate reporting. It is about creating the conditions where AI can operate effectively across the organization. A mature PPM strategy helps organizations move from fragmented project data toward governed operational intelligence. That foundation becomes increasingly valuable as AI adoption expands across portfolio planning, forecasting, risk management, resource optimization and strategic investment decision making.
AI success starts long before the model
There is growing consensus across industry that successful AI adoption depends less on the sophistication of the AI itself and more on the quality of the surrounding operational environment.
AI does not replace the need for structure, governance or delivery discipline. In many cases, it increases their importance. Organizations that treat data governance, operational maturity and portfolio visibility as strategic priorities are likely to be far better positioned to realize long term AI value. Those that do not may continue to struggle moving beyond disconnected pilots and isolated experimentation.
The future of enterprise AI will not be built solely on algorithms. It will be built on trusted operational foundations.
About the author
:format(webp))
David Joynes is an account director of large aerospace and defense enterprise at Cora Systems. He leads strategic relationships and drives growth with the world’s largest A&D companies.
:format(webp))
:format(webp))
:format(webp))