May 3, 2026
The next wave of value creation
Despite a 200% rise in GenAI spending, most pilots stall. We explore why scaling remains elusive.
Across enterprise AI deployments, one pattern is becoming clear: the biggest obstacle isn’t the technology. It’s everything around it.
While organizations continue investing heavily in generative AI – spending rose an estimated 200% in 2025 alone – many initiatives still struggle to move beyond pilot stage. The difference between those that scale and those that stall often comes down to a single overlooked factor.
We spoke with Klaus Gorges, Director of Global Solution Design, about what he’s seeing across real implementations – and what actually determines whether GenAI delivers operational value.
K: Most organizations understand where they want to go with AI – strategy is rarely the issue. The real challenge starts when they try to operationalize it. GenAI performs best when it’s deployed on top of standardized, well structured, high volume processes. But in reality, very few companies have achieved that level of process maturity.
When GenAI meets fragmented workflows, it doesn’t fix the gaps – it amplifies them. That’s why many projects move fast in pilot mode but stall during scaling.
The models are advancing incredibly quickly. What slows organizations down is everything around the model – operating models, governance, infrastructure, and skills. In regulated environments especially, the absence of strong governance structures becomes a real blocker.
Successful execution requires orchestration across processes, controls, and people, not just technology.
In our experience, three areas create the biggest friction.
First, processes are rarely industrialized. GenAI thrives on clear inputs and outputs, but many workflows are inconsistent or undocumented.
Second, integration challenges. Legacy systems, siloed data, and fragmented IT landscapes slow down deployment.
Third, skills shortages – 94% of leaders report gaps in AI critical capabilities. These issues rarely surface in a controlled pilot but become painfully visible when scaling.
It’s foundational. Around 80% of enterprise data is unstructured – emails, documents, PDFs – which makes it difficult to use effectively without advanced processing capabilities.
Organizations often underestimate how much work is required to prepare data for AI. But once that foundation is in place, the impact can be significant. For example, automation of repetitive tasks can free up 60–70% of employee time, allowing teams to focus on higher-value work.
Human in the loop validation is essential. AI outputs can contain inaccuracies or bias, so robust governance and quality assurance mechanisms must be integrated into the workflow.
The goal isn’t full autonomy – it’s controlled autonomy. AI handles volume; people handle exceptions and final decisions.
The goal isn’t full autonomy – it’s controlled autonomy. AI handles volume; people handle exceptions and final decisions.Klaus Gorges Director of Global Solution Design, SPS
When speed, compliance, and scale matter. Building enterprise-grade AI requires infrastructure, monitoring, validation frameworks, and governance controls. That’s a significant investment to develop from scratch.
Organizations that partner with experienced providers can often leapfrog the build phase and move directly to measurable outcomes. In many cases, it’s the fastest path from experimentation to value.
Winning in times of Disruption
Learn how organizations move from AI pilots to scalable impact through agentic automation, human-in-the-loop models and intelligent automation designed for real business processes.
Successful organizations treat AI as an operational transformation, not a technology initiative. They focus on process maturity, governance, and execution discipline.
They also recognize that scaling AI is not a single project. It’s an ongoing capability. That’s why operating models, monitoring mechanisms, and clear ownership structures matter just as much as algorithms.
We’re moving toward AI systems that can plan, decide, and execute multi step tasks with minimal human intervention – what we refer to as agentic automation.
As autonomy increases, governance becomes even more critical. Organizations will need clear escalation rules, monitoring dashboards, and structured oversight to ensure safe operation.
AI is shifting from “copilot” to “co worker,” and companies with strong operational foundations will benefit most from that transition.
A common misconception is that AI success hinges on selecting the right model. In practice, it depends far more on building the right operational environment around that model.
Organizations that scale GenAI effectively excel in three areas:
As GenAI adoption accelerates, the divide between organizations that experiment and those that achieve sustainable, scalable impact will only grow. The differentiator won’t be strategy—it will be operational readiness.
for expert perspectives on generative AI.
Experienced Solution Designer with a demonstrated history of working in the outsourcing/offshoring industry. Skilled in Business Process Shared Services, Project Implementation and Business Process Outsourcing (BPO).