Skip to content
English

Why AI Projects Fail - and What CIOs Need to Change Now

Artificial intelligence remains the dominant topic on the agenda of many CIOs. Expectations from the boardroom are high, investments continue to rise - and yet results in many organizations fall short.

The latest Gartner CIO Report 1H26 paints a remarkably clear picture: organizations are investing heavily in AI, but increasingly struggle to turn pilot projects into real business value. According to Gartner, 59% of all AI initiatives never reach production. At the same time, 83% of CEOs are increasing their AI investments.

The key takeaway: the problem is often not the AI itself.  The problem is the lack of adaptability in existing enterprise systems.

 

AI exposes the limits of outdated enterprise software

In many organizations, modern AI initiatives run straight into an operational reality that has been built up over years or decades:

  • monolithic core systems
  • hard-coded business logic
  • isolated data structures
  • complex custom software
  • fragmented process landscapes
  • high manual integration overhead

As long as these structures remain in place, every new technology - including AI - will inevitably get stuck on organizational and technical friction.

That is exactly what the Gartner report reflects. Gartner calls on CIOs to:

  • replace fragmented AI pilots with unified platform strategies,
  • modernize architectures,
  • reduce technical debt,
  • streamline governance
  • and continuously evolve AI as a 'living product'.

Notably, the Gartner report says relatively little about AI models themselves. Instead, the focus is on platform consolidation, governance, modernization, and technical debt.

This makes one thing clear: the real challenge is not the availability of AI technology, but the adaptability of operational systems.

 

AI amplifies architectural weaknesses

A common mistake is treating AI as a standalone technology initiative.

In reality, AI acts more like an accelerator of existing structures: well-positioned organizations move faster. Rigid organizations become visibly slower.

Because AI increases:

  • the pace of change,
  • the need for real-time data,
  • the number of automated decisions,
  • integration requirements,
  • and the need for continuous process adjustments.

This dramatically raises the demands on the underlying enterprise architecture.

Technical debt does not shrink as a result, it becomes strategically more relevant.

Gartner puts this indirectly but clearly: CIOs must modernize their architectures and reduce technical debt to meet future demands.

In the AI era, legacy structures are becoming an increasingly significant barrier to innovation.

 

The new core competency: adaptability

For many years, topics like digitization, cloud migration, and efficiency improvement dominated the agenda.

Today, the priority is shifting.

Successful organizations need to be able to continuously adapt their processes, applications, and business models - faster than market shifts, regulatory changes, or new technological possibilities emerge.

Adaptability is becoming a strategic core competency.

The central question is no longer: "How do we deploy AI?"
But: "How quickly can we change our organization when AI creates new opportunities?"

Gartner underlines this. CIOs must not only drive individual AI initiatives, but simultaneously:

  • reduce technical debt,
  • consolidate fragmented platforms,
  • make governance scalable
  • and modernize architectures to deploy AI productively and sustainably.

This shifts the focus from individual AI projects to the ability of the entire organization to continuously adapt.

 

Why classic ERP and legacy modernization often hits a wall

The challenge is particularly visible in highly customized ERP and legacy systems - such as Microsoft Dynamics NAV or RPG applications on IBM i.

Many organizations have developed these systems over decades, together with partners and internal teams, into business-critical solutions. A significant share of their operational process logic and organization-specific knowledge is embedded in these applications.

But that is precisely what creates growing challenges:

  • increasing technical debt,
  • complex upgrade projects,
  • rising integration requirements,
  • heavy dependence on specialist knowledge
  • and a declining ability to adapt the existing architecture.

In the RPG space, a shortage of qualified professionals is adding further pressure. At the same time, many of these applications remain critical for production, logistics, or ERP processes.

Many organizations respond with classic 'rip-and-replace' strategies - fully replacing existing core systems with new standard platforms.

But this approach often comes with significant risks:

  • multi-year transformation projects,
  • re-implementation of core processes,
  • loss of custom business logic
  • and considerable disruption to day-to-day operations.

Modern modernization no longer means starting over

The Gartner CIO Report makes clear that CIOs today are primarily focused on continuous adaptability, scalable governance, and modern platform architectures.

That is why modernization approaches are gaining ground that do not discard existing business logic, but instead develop it further in a structured way.

Modern modernization no longer means fully replacing existing systems.

Instead, more and more organizations are choosing model-driven software development, in which existing business logic is analyzed, abstracted, and gradually migrated to modern architectures.

The result is a software landscape that is not only technologically modernized, but remains adaptable in the long term.

Particularly when modernizing Dynamics NAV or RPG applications on IBM i, the question is less and less about replacing proven processes with standard software. Instead, the focus shifts to:

How do you migrate existing business logic into a modern, long-term adaptable architecture?

Model-driven platform approaches enable step-by-step modernization, in which existing data structures and process logic are analyzed and migrated - without a big-bang migration and without losing existing customization.

 

Modernization becomes business strategy

The Gartner report is clear: modernization is no longer a purely IT topic.

It is becoming a strategic factor for:

  • innovation capability,
  • resilience,
  • governance,
  • cost efficiency,
  • cybersecurity
  • and AI scalability.

CIOs are therefore facing a fundamental shift in their role: they are no longer solely responsible for technology. They create the structural conditions that allow the organization to respond to change.

Organizations today do not need more isolated AI pilots. They need operational systems that can structurally support change.

That is precisely why modernizing core systems is increasingly becoming a strategic prerequisite for successful AI initiatives.

 

Conclusion

The current AI hype tempts many organizations to introduce new tools and models as quickly as possible.

But the real bottlenecks often run deeper: in the operational core systems of the organization.

As long as applications remain difficult to adapt, processes stay fragmented, and technical debt keeps growing, AI will not be able to fully realize its potential either.

The defining competency in the AI era is therefore not automation alone. It is adaptability.

Organizations that continuously modernize and keep their software landscape flexible create the foundation to implement innovations faster, manage risks more effectively, and sustainably translate AI into added value.

Because successful AI does not emerge from rigid system landscapes.

It emerges from modern enterprise software that enables continuous change.