The AI infrastructure boom is usually framed around hyperscalers and GPUs. But the more fundamental question is: who is actually building—and controlling—the systems that make AI possible? Global operators like NTT are no longer just delivering space and power. They’re assembling integrated platforms that span data centers, connectivity, edge, and enterprise transformation. At the same time, hyperscale demand is converging with enterprise AI adoption, reshaping how infrastructure is specified, financed, and deployed. This opening fireside looks at where control of the AI stack is consolidating—and where it remains fragmented.

Drawing on NTT’s global expansion, hyperscale partnerships, and edge initiatives, along with Miller’s long-view industry perspective, the conversation explores:

• How large-scale operators are evolving from colocation providers into full-stack infrastructure platforms

• How hyperscale and enterprise AI demand are converging in practice • What it takes to deliver capacity globally at scale in a power- and grid-constrained environment

This session sets the tone for the Summit with a grounded look at how AI infrastructure is actually being built, who is shaping it, and which models are likely to endure.