Google invests $15.1 billion in Finland AI infrastructure
The tech giant's largest European investment to date will leverage nuclear power to sustain high-compute AI workloads.
Google has announced a record-breaking €13 billion (approximately $15.1 billion) investment to expand its AI and digital infrastructure in Finland. The move marks the company's largest single investment in Europe to date, signaling a massive bet on the Nordic region to anchor its global AI compute capabilities.
The buildout, scheduled for 2027 and 2028, centers on the opening of data center operations across three new locations: Kajaani, Muhos, and Vaala. To meet the immense power requirements of large language models and AI processing, Google has signed a 22-year agreement to support the life extension of the Loviisa nuclear power plant, under which the company will purchase 50% of the plant's energy output.
The Nordic Advantage
Finland has emerged as a primary hub for AI infrastructure due to a combination of geographic and systemic advantages. The country's naturally cold climate significantly reduces the energy and costs associated with cooling massive server farms. More critically, Finland offers access to stable, carbon-free power, which is essential for tech giants facing increasing pressure to meet sustainability goals while scaling energy-hungry AI operations.
Solving the Energy Problem
This strategic shift underscores the growing "energy problem" facing the AI industry. As the race for more powerful models intensifies, the demand for electricity has outpaced the capacity of traditional renewable grids. By explicitly linking its infrastructure expansion to nuclear energy, Google is prioritizing baseload power—energy that is constant and reliable—over intermittent sources. This move highlights the critical role nuclear power is expected to play in the future of big tech's infrastructure strategy.
Industry Implications
Google's commitment suggests a broader trend of diversifying compute hubs away from traditional tech centers toward regions that can provide sustainable, high-density power. The scale of the Loviisa agreement demonstrates that the AI race is no longer just about chip architecture or algorithmic efficiency, but about securing the physical energy supply chains required to run them.
Industry observers will now watch for similar moves from other hyperscalers in the Nordic region. While the infrastructure timeline is set for 2027 and 2028, the immediate impact is a clear signal that nuclear energy is now a cornerstone of the AI infrastructure playbook.