Governments worldwide compete fiercely to lead the AI revolution, rolling out policies, funding initiatives, and national strategies to claim a slice of this transformative industry. The UK launched a comprehensive plan in March to develop the infrastructure essential for AI innovation and economic expansion. The US advances an action plan targeting global leadership. Japan’s AI Promotion Act, active since September 2025, positions the nation as the world’s most welcoming destination for AI development.
However, many initiatives overlook a critical reality: AI progress hinges on physical infrastructure rather than just software or policy goals. Training and deploying large-scale models requires vast compute resources, which depend on available land, reliable power, and rapid deployment timelines. Investors now prioritize locations with operational, cost-effective infrastructure over ambitious announcements.
To secure sustained AI investment, governments must prioritize practical enablers like land availability, energy supply, and compute capacity.
Treat Energy as Core AI Infrastructure
Energy forms the bedrock of AI operations. Governments should integrate power generation, grid connectivity, and data center expansion into a unified strategy for swift, dependable, and affordable rollout. AI tasks demand massive energy, sensitive to costs and supply. Projections indicate AI data centers will consume 8-12% of US electricity by 2030, up from 3-4% currently.
When energy planning, grid access, and data centers operate in silos, projects grind to a halt. Recent UK research reveals that 24.5% of tech firms allocate over a third of their AI budgets to energy costs, while 37% cite price fluctuations as a major barrier to AI advancement. Reliable, economical energy draws investment; its absence pushes capital elsewhere.
Prioritize Speed in AI Deployment
In AI infrastructure, delays undermine projects. Rapid hardware cycles mean 12-18 month setbacks can invalidate original plans. Firms often secure GPUs but face hurdles in permitting, planning, and grid connections that prevent production readiness.
Ireland exemplifies this: surging data center demand strained Dublin’s grid, prompting pauses on new connections in key areas. Even with land and funding secured, projects waited years for power. Governments streamlining approvals and accelerating activation gain a decisive edge—access alone no longer suffices; operational speed does.
Focus on Compute Access Over Innovation Funding
While funding AI research and startups remains vital, scaling demands compute availability. Europe’s Horizon Europe invests billions in AI, yet high-performance compute stays fragmented, lagging US-scale facilities. Many European AI firms turn to overseas clouds for growth, balancing sovereignty against economics.
In the UK, 45% of organizations factor data sovereignty into strategies, but 43% prioritize cost and performance. Local infrastructure must deliver capacity affordably and promptly, or workloads migrate.
Promote Distributed Regional Infrastructure
AI facilities cannot cluster in few spots; success depends on land, power, and grid realities. Scaling fastest occurs in regions primed with these assets, avoiding costly new builds. Distributed models ease grid strain, boost resilience, and enable broader growth.
Bournemouth, UK, hosts viable data centers. Eastern Europe rises as a hub—Romania draws interest with over half its energy from hydro and nuclear, costs up to 40% below Western Europe, and a thriving AI scene featuring 146 companies, including unicorn UiPath valued at $35 billion.
Integrate Sovereignty, Economics, and Skills
Sovereignty demands, cost controls, and talent gaps intersect at deployment. Regulated sectors need local compute that’s viable and backed by skilled teams. Shortages extend beyond AI developers to infrastructure experts in power, cooling, networking, and operations.
Governments must develop these skills alongside infrastructure to ensure stable, secure environments. Without them, investments fail to yield lasting impact.
Execution Drives AI Investment
Success boils down to delivery. Nations providing scalable, real-world infrastructure—aligning energy, planning, compute, and talent—will capture AI capital. Others risk relying on foreign AI rather than shaping it.
