Key Takeaways
⇨ Many businesses are enthusiastic about AI and believe they have the skills to scale it, but a large number are still struggling to move beyond early-stage deployments.
⇨ Successful AI adoption depends on more than technology. Stable ERP processes, reliable data, governance, and clear business objectives are becoming essential foundations.
⇨ CIOs and ERP leaders are under increasing pressure to build stronger data governance, develop AI capabilities across teams, and establish clear measures for ROI.
Organisations are moving quickly to experiment with artificial intelligence, but many are finding that turning promising pilots into large-scale business capabilities is much harder than expected.
Infor’s Enterprise AI Adoption Impact Index highlights a growing disconnect between how confident companies feel about their AI capabilities and how successfully they are actually deploying AI across their operations.
The research, based on responses from 1,000 decision-makers across the United States, United Kingdom, Germany, and France, found that many organisations remain stuck in the early stages of AI adoption.
What makes the findings particularly interesting is that while a significant number of businesses struggle to scale AI, around 80 percent of respondents believe their organisation already has the internal capability to implement it.
For CIOs, ERP leaders, and transformation teams, this points to a familiar problem.
The challenge is often not getting an AI pilot to work. The real challenge is integrating AI into everyday business processes in a way that is secure, governed, measurable, and scalable.
AI Ambition Is Moving Faster Than Operational Readiness
The index focuses less on whether organisations are experimenting with AI and more on whether those experiments are creating meaningful operational impact.
According to the findings, nearly half of organisations remain in the early stages of deployment, despite having already developed proofs of concept or launched initial AI initiatives.
This is the stage many businesses now describe as “pilot purgatory.”
AI experiments may generate excitement and demonstrate potential, but they never become deeply integrated into the workflows that employees use every day.
There are several reasons behind this.
Data security was identified as the biggest challenge, followed by a shortage of AI talent and uncertainty around return on investment.
These issues highlight why AI adoption is increasingly becoming an enterprise transformation challenge rather than simply an IT project.
Technology leaders now need to answer difficult questions around data ownership, security, governance, skills, accountability, and business value before AI can move into core operational systems.
Infor’s findings suggest that many organisations are running multiple AI experiments without connecting them to end-to-end processes, consistent data models, or operating structures that would allow them to scale.
The result can be a growing collection of AI tools that create more complexity without delivering meaningful improvements in productivity, customer experience, or operational performance.
What This Means for ERP Leaders
AI cannot scale effectively on top of unstable business processes.
For ERP and transformation teams, AI readiness starts with the fundamentals.
Organisations need reliable processes, strong data quality, clear governance, and defined responsibilities before introducing AI into critical workflows.
The focus should not simply be on launching more pilots. It should be on identifying where AI can become part of an existing business process and deliver a measurable improvement.
Where Embedded AI Is Already Delivering Results
The difference between experimentation and real value often comes down to how closely AI is connected to operational workflows.
Infor points to examples where AI has been applied to specific business problems rather than being deployed as a general-purpose technology experiment.
In one example, warehouse optimisation helped reduce employee travel distance by 25 percent. In another, automation improved order entry processes and allowed employees to spend more time on higher-value work.
These examples are important because they demonstrate what successful enterprise AI often looks like.
It is not necessarily a dramatic transformation overnight.
Instead, value can come from solving specific operational problems, reducing repetitive work, improving decision-making, and making existing processes more efficient.
When AI is connected directly to ERP, warehouse, supply chain, finance, or workforce workflows, it becomes easier to measure whether it is actually creating business value.
Infor Focuses on Moving AI From Pilots to Execution
Infor is responding to these challenges with new capabilities across its Velocity Suite and an enhanced version of its Agentic Orchestrator.
The goal is to help organisations move beyond isolated AI experiments and toward more structured, governed automation.
Velocity Suite brings together tools, accelerators, and industry-focused approaches designed to help customers deploy and modernise Infor’s cloud solutions more efficiently.
For ERP teams, this could help address an important problem.
AI initiatives are much harder to scale when they are introduced on top of fragmented systems, inconsistent data, or poorly integrated processes.
Creating a more standardised foundation for migration, configuration, and integration can make it easier to introduce AI into business workflows later.
The enhanced Agentic Orchestrator takes this a step further by focusing on how multiple AI agents can work across different business functions.
These agents can support workflows across areas such as supply chain, finance, and workforce management while operating within defined governance and security frameworks.
Infor says AI-driven actions can be logged, policies can be enforced, and human oversight can remain in place for higher-impact decisions.
This is becoming increasingly important as enterprise AI moves beyond simply answering questions.
The next generation of AI systems is expected to recommend actions, trigger workflows, handle exceptions, and potentially complete tasks with varying levels of autonomy.
What This Means for ERP Leaders
The next major AI battleground will be orchestration.
Businesses are unlikely to rely on a single AI tool to manage every process.
Instead, they may increasingly use multiple specialised AI agents working across different business functions.
The challenge will be managing how those agents interact with data, systems, workflows, and employees.
Platforms that can provide strong orchestration, security, logging, governance, and human oversight could therefore become increasingly valuable.
ERP Readiness Will Become AI Readiness
One of the most important messages from the Infor index is that organisations should assess their operational readiness before scaling AI.
An organisation may have talented employees and access to powerful AI technology, but that does not automatically mean it is ready to deploy AI across critical business processes.
ERP and architecture teams will increasingly need to evaluate areas such as:
- Data quality and accessibility
- Process consistency and stability
- Security and governance
- Integration across business systems
- AI skills across technical and business teams
- Human oversight and accountability
- Clear measures of business value
These areas will become essential parts of enterprise AI roadmaps.
For organisations running complex environments involving multiple ERP systems, cloud platforms, supply chain applications, and HR systems, the challenge becomes even greater.
AI will need to operate across connected processes rather than within isolated applications.
That will require much closer collaboration between ERP teams, data leaders, IT departments, business units, and employees who actually use these systems every day.
Measuring ROI Will Separate Real AI Programs From Hype
One of the biggest barriers identified in the research was uncertainty around ROI.
This may become one of the most important factors separating successful AI programs from endless experimentation.
Organisations that define success only after deploying AI will struggle to determine whether an initiative has actually worked.
Instead, transformation leaders need to establish measurable outcomes before scaling.
Those metrics could include:
- Reduced processing time
- Lower operational costs
- Faster order fulfillment
- Improved inventory accuracy
- Fewer manual tasks
- Reduced employee workload
- Faster decision-making
- Better customer service
The most successful AI initiatives will likely be the ones that connect technology directly to a measurable business outcome.
What This Means for ERP Leaders
Measurement discipline will be critical to scaling AI successfully.
AI enthusiasm alone will not create business value.
The organisations that define clear objectives, establish baseline performance, and measure outcomes consistently will be in a stronger position to decide which AI initiatives deserve further investment.
The Bigger Challenge Is Not AI Technology
The Infor index highlights an important reality for enterprise leaders.
The biggest obstacle to AI adoption may not be choosing the right model or finding the latest technology.
The bigger challenge is creating an environment where AI can operate safely and effectively within real business processes.
For ERP leaders, that means the foundation matters more than ever.
Strong data governance, stable processes, reliable integrations, clear accountability, and measurable outcomes are becoming prerequisites for successful AI adoption.
The companies that move beyond pilot purgatory will not necessarily be the ones experimenting with the most AI tools.
They will be the ones that successfully embed AI into the way their business already operates and can clearly demonstrate the value it creates.
As AI becomes more deeply connected to ERP, supply chain, finance, and workforce systems, the conversation is likely to shift from “What can AI do?” to a much more important question:
“Where can AI create measurable value, and are our systems ready to support it?”
