The AI gold rush is rewriting the rules of energy, politics and society at once. Europe's answer so far has been a sovereignty debate. Our latest newsletter argues the bigger question is what kind of society AI will create.
Developments in AI are running at breakneck speed. Three headlines stand out:
- Anthropic files to go public in trillion-dollar debut
- AI Data Centers Are Sending Power Bills Soaring
- California city votes to permanently ban datacenters
Three different story lines are unfolding at once.
First, there is the mindboggling scale of financing. Open AI and Anthropic are debuting soon on public markets with valuations close to $1 trillion each. Everyone is piling in. Goldman Sachs estimates that Amazon, Microsoft, Google and Meta alone will spend $5.3 trillion on AI infrastructure by 2030. Some financiers already describe the AI buildout as the largest investment cycle in the history of capitalism.
This is a veritable race, with the US leading the pack and China following closely. The AI world is largely organised around two centres of gravity. The US West Coast and the Chinese East Coast. Europe is barely on the map. But it’s trying to catch up (note Softbank’s announcement of a €75 billion data centre investment in France).
But the sheer expansion is not without consequences. Enter headline number two.
AI with its data centres is incredibly energy and water intensive. In the US, power prices in some regions have jumped by 76% due to rampant data centre demand. In Ireland, studies are concluding that data centre growth has added roughly €360 to the average household electricity bill over the years. In Europe, McKinsey estimates that data centre electricity demand could more than triple by 2030.
As such, the expansion of data centres in many places is becoming a hot political issue with some cities and regions considering bans. Like every gold rush, the AI boom is creating enormous fortunes. But it is also creating new dependencies, new environmental pressures and new political questions that societies are only beginning to confront.
But that’s not the main debate in Europe.
Here, it is by and large a sovereignty debate. How do we catch up? How do we scale? How do we reduce our dependencies?
These are legitimate concerns. Europe cannot afford to become a spectator in one of the most consequential technological transformations in history. And the European Commission has rightly now published its Tech Sovereignty Package.
Yet there is a risk in viewing AI only through the prism of competition. AI influences what information people see. It changes how children learn. It alters labour markets, creative industries and education systems. It is becoming a major factor in electricity demand, water consumption and public infrastructure planning.
The effects are manifold.
The question is therefore not only whether Europe can build AI. The question is what kind of society widespread AI adoption will create.
This is exactly why we have partnered with the Green European Foundation to launch a Knowledge Community on Tech & Democracy. We want to look at the issue comprehensively. Over the coming months, participants will explore some of the defining political questions raised by AI and digital technologies: technological sovereignty, democratic resilience, social media and its impact on children, the environmental footprint of AI, and the changing relationship between technology, power and democracy.
In the meantime, I invite you to join a debate next week on the promises and risks of digital sovereignty, organised by our Human Security Unit in Vienna, read our interview with Rebecca Lenhard, spokesperson for digital policy for the Greens in the Bundestag, and save the date of 30 June in the late afternoon, for an event on the EU’s Cybersecurity Act, which we are co-hosting with the German Association of the Electrical and Digital Industry (ZVEI).
Read more in our Böll EU 05/2026 Newsletter!
Warm regards,
Roderick Kefferpütz
