Sovereignty is shifting from compliance issue to core risk strategy
AI is transformative—but only as good as the data behind it
Enterprises must balance innovation with resilience in a rapidly changing landscape
In a world of fracturing supply chains, rising geopolitical tension and accelerating AI adoption, enterprises are being forced to rethink how they design their systems.
Logan Wolfe, Partner for Global Enterprise Transformation, AI and Tech Strategy at Kyndryl, argued that sovereignty is no longer a technical afterthought—it is a foundational design principle. In an interview, he outlined why businesses need to rethink risk, data and infrastructure in an age where uncertainty is the default.
For Wolfe, the future of enterprise infrastructure isn’t about choosing between sovereignty and innovation—it’s about designing systems that can deliver both.
In a world where uncertainty is constant and technology evolves by the hour, the winners will not be those who move fastest, but those who build systems resilient enough to keep up.
Sovereignty is a fundamental risk framework
Steve Saunders: Kyndryl sits in a really interesting position around enterprise and industrial digitalization. What’s your perspective on what matters in this space?
Logan Wolfe: A lot of people still see this as a technical issue—an IT issue. But if you zoom out, we’re living in a world of increasing geopolitical tension and fracturing supply chains.
People tend to look at sovereignty and AI as technology problems, or maybe regulatory compliance problems. Those are drivers, but we frame sovereignty differently, as a fundamental risk
mitigation approach.
Uncertainty is part of the course now. You need to design for sovereignty as a strategy for optionality. Technology is just the tool to get that done.
Saunders: So you’re flipping the script, treating sovereignty as a risk layer to design around. Is there also a commercial upside?
Wolfe: It’s a service area like any other, but we look at sovereignty across three domains. First is data—the oil of the digital economy. If you know how to refine it and use it, it underpins everything.
Second is operations—the ability to continue operating independently of undue external influence. Third is technology and supply chains—using systems that are resilient to surveillance or disruption from actors you don’t trust.
Across those domains, sovereignty underpins competitiveness and profitability.
But it’s not something you can just buy as a product, as a SKU. It comes with trade-offs. You might try to detach from hyperscalers, but then you lose access to innovation. So it’s always a balance.
Saunders: Enterprises are being pulled toward hyperscalers owning the full stack. Is that a risk?
Wolfe: It depends on the use case and the organization’s appetite for sovereignty.
Take AI. You might deploy a custom model on AWS to stay competitive. But you can also set up failover pipelines to a local stack that replicates that capability. That way you stay innovative while building resilience.
It’s a palette. You pick and choose providers and architectures based on your risk tolerance. That’s where we come in, helping customers design those systems.
AI is only as good as the data it’s fed
Saunders: There’s a lot of hype around AI. Are we over-rotated?
Wolfe: It’s not everything people claim, but it is transformative. We see it everywhere: predictive maintenance, fraud detection, customer support, security operations. But the reality is simple: AI is only as good as the data it’s fed. If your data is inconsistent, incomplete or stuck in legacy systems, it throttles everything.
Organizations try to speed-run AI deployment, but it’s a gradual process. That’s why you see studies suggesting most AI use cases don’t deliver ROI.
Saunders: So step one is getting the data right?
Wolfe: Exactly. It’s a data transformation process. You consolidate, clean, label and vectorize the data—basically make it AI-ready. Then you end up with a unified data lake that can support multiple use cases. That’s the foundation. Without it, nothing scales.
Saunders: Enterprises struggle to scale successful AI projects across the organization.
Wolfe: That’s a big part of what we do. It’s not just about building something once, it’s about running it continuously. The world is changing fast. Data changes. Regulations change. The EU AI Act, GDPR updates—everything evolves. You need systems that can adapt, with sovereignty and compliance baked in from the start.
That means data engineering, explainability, governance. It’s a whole architecture.
Saunders: There’s so much happening—agentic AI, tokenization, new paradigms every week. How should enterprises respond?
Wolfe: Even for experts, keeping up is a full-time job now. Organizations need to take a programmatic approach. You don’t have to be first. You can be in the top 10% of adopters and still lead. The key question is: are you solving real problems, or just chasing hype? That’s what business leaders struggle with.
Speed vs responsibility is the real divide
Saunders: Is Europe better than the U.S. at all of this?
Wolfe: “Better” is subjective. The U.S. is driving massive data center build-outs—that creates jobs and infrastructure, but also raises energy concerns. Europe tends to lead on regulation and ethics. Some see that as slowing progress; others see it as responsible governance.
Ultimately, it comes down to goals. Is speed of innovation more important, or responsible deployment?
AI is just a tool. The real question is what outcome you’re trying to achieve.