Cloud 2.0: From Infrastructure to Insurance Intelligence

On February 13, insurance leaders gathered for Cloud 2.0: The Architecture of Insurance Intelligence, a session hosted with Qorus in partnership with NTT DATA. The conversation which I had the opportunity to lead, alongside industry executives Nicki Tang, CIO of Tokio Marine HCC International, and Dipu KV, Senior President at Bajaj General Insurance, quickly moved beyond familiar cloud talking points. 

This wasn’t another discussion about migration roadmaps or hosting strategies. It was something far more strategic - and, for many organizations, more uncomfortable. 

This wasn’t another discussion about migration roadmaps or hosting strategies. It was something far more strategic — and, for many organizations, more uncomfortable.

Because the uncomfortable truth is this: cloud is no longer a differentiator.

A decade ago, moving to the cloud signaled innovation. Today, it’s table stakes. When more than 90% of insurers already run core workloads in cloud environments, simply “being in the cloud” doesn’t make you modern. It just means you’ve caught up.

So the question facing executives has quietly changed. It’s no longer why cloud, or even how fast can we migrate. It’s how do we turn cloud into intelligence?

That shift — from infrastructure to intelligence — defined the tone of the session from the outset. Cloud has evolved from being the place where applications live into something closer to an operating system for enterprise reinvention. In practice, that means cloud only creates value when it enables AI, better decisions, and faster change across the enterprise.

This is where many insurers hit a wall. They completed their migrations, modernized hosting, reduced some costs… and then wondered why nothing fundamentally changed. Products weren’t shipping faster. AI pilots weren’t scaling. Data still felt fragmented. Innovation still felt heavy.

Because moving infrastructure without rewiring the enterprise rarely delivers transformation.

What stood out in the discussion was how pragmatic the leading insurers have become. Rather than chasing heroic, multi-year “rip and replace” programs, they’re focusing on architectural moves that unlock speed now. As Nicki Tang described from her experience at Tokio Marine HCC, the breakthrough often comes not from replacing every legacy system, but from decoupling them — introducing modern integration layers that free digital and data capabilities to evolve independently. It’s a shift from perfection to momentum: modernize what matters first, and don’t wait for a pristine estate before starting with AI.

At the same time, the data conversation has matured. Scaling AI isn’t about accumulating more data, but about trusting the right data. Ownership, governance, and clarity of purpose matter more than volume. Data increasingly behaves less like an IT by-product and more like a product with accountability, standards, and business outcomes attached. Without that discipline, AI simply automates inconsistency.

The cloud debate itself has also grown more nuanced. The early “cloud-first at all costs” mindset has given way to deliberate hybrid strategies. Public cloud for agility and experimentation. Private or sovereign environments for control, compliance, and predictable economics. It’s no longer about ideology; it’s about workload placement discipline. The right workload, on the right platform, for the right reason.

Perhaps the most candid shift came around cost and accountability. AI experimentation is exciting, but unchecked it becomes expensive very quickly. As Dipu KV emphasized, every initiative must show measurable value, whether through productivity gains, financial returns, or clear strategic impact. Projects without a defensible business case simply don’t progress. It’s a tough filter, but it’s also what allows transformation to scale without losing the CFO’s confidence. In that sense, FinOps isn’t a constraint on innovation — it’s what makes sustainable innovation possible.

Even the industry’s newest frontier, agentic AI, is being approached with caution. While the vision of autonomous agents is compelling, insurance is not an environment where hallucinations are tolerable. Decisions carry real financial and regulatory consequences. So the path forward looks measured: tightly bounded use cases, trusted data sources, strong auditability, and humans firmly in the loop. Ambition remains high, but execution is deliberately staged.

Stepping back, a bigger pattern emerges. The real competitive advantage isn’t any specific AI model. Those are quickly becoming commodities. Everyone has access to the same tools and platforms. What separates leaders is the architecture around them — how quickly they can adapt, how confidently they can trust their data, how predictably they can manage costs, and how safely they can scale intelligence.

The differentiator isn’t AI itself. It’s the ability to operationalize intelligence at scale.

That’s why Cloud 2.0 feels less like a technology upgrade and more like a leadership test.

It forces a choice. Continue optimizing infrastructure, or design an enterprise that learns and adapts faster than competitors.

Cloud 1.0 made insurers digital.

Cloud 2.0 makes them intelligent.

And in a market where products are copied quickly and margins are constantly pressured, intelligence — not infrastructure — may be the only edge that lasts.