Google has unveiled one of Europe’s largest artificial-intelligence infrastructure investments, committing €13 billion, approximately $15.1 billion, to Finland over the next two years.
The program will expand Google’s digital infrastructure and supporting energy systems across Hamina, Kajaani, Muhos and Vaala, as demand continues rising for services including Gemini, Search and other cloud-based products. Google described the commitment as its largest single investment in Europe. (blog.google)
But the strategic significance of the announcement extends well beyond new data centers.
At the center of the program is a long-term energy agreement with Finnish utility Fortum designed to secure large volumes of reliable electricity from the Loviisa nuclear power plant.
Fortum and Google have signed a 22-year power purchase agreement covering up to 50% of Loviisa’s generating capacity, helping provide the revenue certainty needed to extend operations at the nuclear facility through 2050. (Fortum)
Google said this is its first nuclear-energy agreement outside the United States. (Reuters)
The structure represents a significant evolution in the relationship between hyperscale technology companies and power infrastructure.
Rather than simply purchasing electricity from an existing grid, major AI operators are increasingly helping underwrite the long-term economics of the generation assets required to support their computing demand.
AI Is Becoming an Energy Infrastructure Business
The rapid expansion of generative AI has transformed electricity availability into one of the technology sector’s most important strategic constraints.
Large AI data centers require enormous quantities of power not only to operate processors, but also to run networking equipment, storage systems and cooling infrastructure.
Google’s Finland strategy therefore combines computing infrastructure with multiple layers of energy investment.
In addition to the nuclear agreement, Google said it will add new onshore wind capacity and contract a 94-megawatt battery storage system designed to support grid stability during periods of low wind and high demand. (blog.google)
Fortum and Google have also signed a memorandum of understanding to explore additional nuclear, renewable-energy and flexibility projects in Finland. (Fortum)
The companies will evaluate potential business models for future nuclear reactors at Loviisa as well as additional locations that could support future Google data centers. (Fortum)
This creates a much broader strategic relationship than a conventional electricity-supply contract.
It links the future of AI infrastructure directly to long-duration power generation, grid resilience and national energy planning.
Nuclear Power Gains a New Economic Driver
For Fortum, the agreement materially improves the economics of extending Loviisa’s operating life.
The two reactors at Loviisa produce approximately 8 terawatt-hours of electricity annually and account for roughly 10% of Finland’s electricity generation, according to Fortum. (Fortum)
Fortum is undertaking an approximately €1 billion investment program to enable the plant to operate through 2050.
The Google agreement will initially cover a smaller volume beginning in 2028 before reaching as much as half of Loviisa’s capacity between 2030 and 2049. (Fortum)
The deal will also support an additional 10 MW power increase, on top of a previously planned 38 MW uprate expected in 2028. (Fortum)
Financial markets reacted immediately.
Fortum shares jumped around 15.5% to 15.8%, reaching their highest level in roughly four and a half years and making the utility one of Europe’s strongest-performing stocks on the day. (Reuters)
The market reaction demonstrates how AI demand is beginning to reprice assets far outside the traditional technology sector.
Power producers, grid operators, nuclear facilities, battery developers and energy-infrastructure companies are increasingly becoming part of the AI investment ecosystem.
Finland Emerges as a Strategic AI Hub
Finland offers several characteristics increasingly valued by hyperscale infrastructure operators.
Its northern climate reduces cooling requirements, while its relatively stable electricity network and substantial low-carbon generation base provide conditions favorable to large computing campuses.
Google has already operated a major data center in Hamina for approximately 15 years.
The new investment dramatically expands that commitment.
Google estimates that construction during 2027 and 2028 will support more than 37,000 jobs nationwide and contribute approximately €3.6 billion annually to Finland’s GDP during the initial construction period. (blog.google)
Once operational, the expanded facilities are expected to support thousands of permanent jobs.
Google is also investing €31 million in local communities and workforce development across the regions hosting its infrastructure. (blog.google)
Europe’s AI Competition Is Becoming Physical
The investment also carries a broader strategic dimension for Europe.
The global AI competition is increasingly moving beyond algorithms and semiconductor design.
Countries now compete for the physical infrastructure required to operate AI at scale:
electricity generation, transmission networks, land, data centers, semiconductor supply chains, cooling systems and connectivity.
Google’s Finland investment illustrates this transition particularly clearly.
The technology company is effectively coordinating capital deployment across computing, nuclear generation, renewable energy, batteries and grid infrastructure within the same expansion program.
That model could increasingly become standard across the hyperscale AI sector.
NEXUS Intelligence View
Confirmed facts: Google has announced €13 billion of investment in Finland over 2027 and 2028, including new digital and AI infrastructure. Google and Fortum have entered a 22-year nuclear power agreement covering up to 50% of the Loviisa plant’s capacity, alongside wind, battery and future energy-development initiatives. Fortum shares rose sharply following the announcement. (Fortum)
NEXUS Intelligence analysis: the most strategically important part of the transaction is not the €13 billion headline.
It is the emerging relationship between AI compute and dedicated energy supply.
The first phase of the AI investment cycle concentrated heavily on GPUs and data-center construction.
The next phase is increasingly about securing enough electricity to keep those facilities operating.
That changes where capital may flow.
Utilities, nuclear plants, transmission infrastructure, energy storage and long-term power contracts are becoming critical components of AI economics.
The Google-Fortum arrangement provides a particularly important model because the technology customer is helping create the commercial certainty required to maintain and potentially expand nuclear generation capacity.
That effectively transforms hyperscale AI demand into a financing mechanism for energy infrastructure.
For governments, the implication is equally significant.
Countries that can combine reliable low-carbon electricity, predictable regulation, available land and strong digital connectivity may attract disproportionately large shares of future AI capital expenditure.
For investors, the AI infrastructure theme should therefore no longer be evaluated only through semiconductor and cloud-company valuations.
The opportunity set increasingly extends across the entire physical stack:
compute → data centers → electricity → grids → nuclear and renewables → storage → cooling → connectivity.
Finland’s €13 billion Google investment may ultimately be remembered less as another European data-center announcement and more as evidence that the global AI race has entered an energy-and-infrastructure phase.