Where hyperscalers and major platforms choose to deploy compute will determine which assets become core infrastructure and which become optional. This panel will explore how AI networks are evolving, with centralised training clusters alongside distributed cloud zones and edge deployments, and what this means for future infrastructure demand.

Understand how hyperscaler roadmaps influence where operators should prioritise capacity expansion, how flexible hyperscalers can be on contract structures, and as deployment accelerates and social license becomes a site-selection and financing variable in its own right, what this means for development planning and long-term financing.

Power availability, permitting timelines, demand visibility and political risk dominate how investors compare European markets, but they don't work independently of each other. Network architecture and access to interconnection help determine where different types of compute can realistically sit. Compare how your peers weigh connectivity alongside power, permitting, demand and political risk in their market selection, discuss which locations currently combine deliverability with a strong network position, and explore how the map may shift as AI inference, sovereignty requirements and network architecture evolve.