AI, energy and geopolitics collide — can decision‑makers steer the “triple transformation”?

A widening gap between capability and capacity

Artificial intelligence is not simply another technology wave; it is reconfiguring the foundations of the global economy. But capability is outpacing the physical and institutional systems that must support it. Data centers, power grids and governance frameworks are evolving slowly. Can infrastructure catch up with algorithms? If not, the gains from AI risk being capped by energy shortages, cooling constraints and fractured rules of the road.

Uneven adoption, tempered expectations

AI is spreading faster across functions than past digital waves, from marketing to R&D, according to a 2024 McKinsey survey. Yet U.S. Bureau of Labor Statistics figures show long‑run productivity gains remain modest — roughly 2% growth in recent years — and many organizations struggle to extract measurable ROI. Gartner finds integration, unclear use cases and model reliability are persistent barriers. It has been reported that current AI tools directly cover only about 5% of business activities, with potential GDP uplift nearer to 1% in the near term, reminding decision‑makers that narrative often outpaces macroeconomic reality.

Energy limits and geopolitical fault lines

Energy and water are fast becoming binding constraints. The International Energy Agency has warned that, by 2030, electricity demand from AI‑related data centres could rival that of a mid‑sized industrial economy, and cooling water stress is already visible in hot, high‑density regions. At the same time, compute, chips and cloud infrastructure are being treated as national security assets: export controls on advanced semiconductors, investments in domestic fabrication, and “sovereign cloud” and data‑localisation policies are proliferating. These moves — and the risk of fragmented regulatory regimes across the U.S., EU and China — increase compliance costs and raise the prospect of uneven access to critical digital resources.

What policymakers and firms must prioritize

The triple transformation is systemic, not siloed. Three actions stand out: embed responsible AI with human‑in‑the‑loop governance, transparency and rigorous model validation from day one; treat digital infrastructure resilience — including grid stability, water for cooling and diversified compute capacity — as a core strategic risk; and prepare for geopolitical fragmentation by tracking export controls, data sovereignty rules and supply‑chain vulnerabilities to preserve operational continuity. The choices made this decade will determine whether AI becomes a durable engine of broad prosperity or a new vector of systemic fragility.