Google just made the largest single investment in its European history. On September 9, the company announced it will commit €13 billion (roughly $15 billion) to AI and digital infrastructure in Finland, and signed a 22-year nuclear power purchase agreement with Finnish energy company Fortum. It is Google’s first nuclear deal outside the United States.

The scale here is not incremental. This is a long-term infrastructure bet that will extend a nuclear plant’s life by two decades, create tens of thousands of jobs, and anchor Finland as a central node in Google’s global AI capacity.

What Was Announced

The investment spans two major components.

First, the data center expansion. Google is building new facilities in Kajaani, Muhos, and Vaala, while also expanding its existing hub at Hamina on Finland’s southern coast. Together, these sites will form a significant cluster of AI compute infrastructure for Google Cloud in Europe.

Second, the nuclear deal. Google signed a power purchase agreement with Fortum covering up to 50% of the output from the Loviisa nuclear plant, a facility that was facing closure as early as 2030. The new agreement extends Loviisa’s operational timeline to 2050. Without this deal, the plant would likely have shut down within four years.

Why Finland

Google has operated a data center in Hamina since 2009, converting a former paper mill into one of its most energy-efficient facilities. Finland offers a combination of factors that are increasingly scarce: stable geopolitics, a cold climate that reduces cooling costs, a reliable electricity grid, and proximity to growing European enterprise customers.

The nuclear agreement specifically addresses what has become the primary bottleneck for hyperscaler expansion: clean, always-on power. Wind and solar cannot provide the consistent baseload that large AI compute clusters require. Nuclear can. Finland’s energy mix is already one of the cleanest in Europe, and locking in 22 years of nuclear capacity essentially removes the energy constraint from Google’s planning horizon in the region.

The Economic Impact

During the construction phase, the investment is projected to contribute an average of €3.6 billion annually to Finland’s GDP and support more than 37,000 jobs. Once the facilities are fully operational, they are expected to sustain around 7,000 jobs annually, with wages averaging 24% above Finland’s median.

For a country of 5.5 million people, these numbers are meaningful. Finland’s government is treating this as a strategic economic partnership, not just a corporate infrastructure deal.

What This Means for Business

For enterprise leaders watching AI infrastructure unfold, this announcement carries a few practical signals.

AI demand is not slowing down. Companies like Google do not commit to 22-year energy contracts unless they have high confidence that AI compute demand will remain strong for decades. The hyperscalers are planning in multi-decade cycles. Businesses that are still treating AI adoption as a wait-and-see decision are operating on a very different timeline to the companies building for them.

Energy is the hidden constraint in the AI race. The Loviisa deal is part of a broader pattern. Microsoft, Amazon, and Google have all signed nuclear agreements in the past year. The constraint on AI scaling is not model quality or hardware. It is reliable, clean power. As these infrastructure investments compound, the cost of running AI services should stabilise over time, but demand for that capacity will grow faster.

European cloud capacity is expanding significantly. For businesses in Europe handling sensitive data under GDPR or the EU AI Act, the growth of local compute infrastructure is relevant. More EU-based data center capacity means more options for compliant AI deployments without routing data outside the region.

The Fortum deal is a template, not an outlier. When a hyperscaler signs a 22-year deal to keep a nuclear plant running, it normalises that model for other markets. Expect more of these agreements to follow, and expect energy producers to start pricing their assets with AI demand in mind.

The Bigger Picture

This investment sits inside a broader wave of hyperscaler infrastructure spending. Across Microsoft, Google, Amazon, and Meta, announced AI infrastructure spending in 2026 has exceeded $300 billion globally. A meaningful portion of that is going into Europe, driven by both regulatory pressure to localise data and genuine demand from European enterprise customers adopting AI at scale.

The Finland announcement is also notable because it signals confidence in existing nuclear, not just the promise of future small modular reactors. Fortum’s Loviisa plant is proven infrastructure. Google is not waiting for next-generation technology. It is paying to keep proven baseload power online because it needs it now.

---

If your business is making AI investment decisions right now, the infrastructure story matters because it tells you something about where the market is heading. The companies building the foundation of the AI economy are making 20-year bets. The question for every business leader is what that means for their own planning horizon.

Enterprise DNA’s advisory services help business leaders build AI strategies that match their actual goals, not a five-year technology cycle. Book a discovery call to talk through how AI fits your business.