Financing Europe’s AI Infrastructure: A New Approach that ensures societal outcomes are met.
AI model providers around the world are loosing money, as the computing costs far outstrip value-creation and thus product prices. Further, AI companies are prioritizing market adoption by using venture- and state-capital to keep prices low and create widespread use, before raising prices to recoup investments. Recouping these massive upfront investments will require a business model that is either subscription-based; OpenAI floated the idea of a 20.000 EUR/month AI Agent; or based on targeted advertising to finance the free or low-cost services being offered.
With the planned Gigafactories initiative, European taxpayers would partially subsidize AI companies' infrastructure investments, helping them achieve profitability and market adoption. However, the current funding approach—which focuses on leveraging public money with private capital—creates a significant risk: companies may ultimately deploy business models (such as targeted advertising or prohibitively expensive subscriptions) that actually limit AI's widespread adoption across Europe.
To address this fundamental challenge, we propose reversing the conventional leverage model. Rather than using public funds primarily to attract private investment, we suggest using public money as low-cost, flexible leverage for private capital through a strategic debt facility.
Thus, the public money should be used to provide a debt facility to companies who are investing in building AI infrastructure. This debt instrument can be structured to be low-interest and with delayed interest payments to make it very attractive for businesses to invest. This would enable taxpayer money to earn rent on the AI infrastructure being built in Europe, creating a return irrespective of the potential productivity gains from AI, and thus hedging the bet on AI for Europe.
Debt can be made attractive and equipped with the constraints required to ensure positive societal outcomes are realized:
Constrain the business models: Apply strict rules on the types of business models that companies taking advantage of the debt-facility may pursue to protect European citizens and industry from over-extraction.
Impact assessments: Companies that utilize the facility must submit an impact assessment that shows the resulting impact on Europe’s key priorities, namely the adoption of AI, productivity increase or sovereignty.
Demand Forecast: To not create an over-supply of infrastructure, any company utilizing the debt-facility must present an adoption plan and associated demand forecast, namely how the infrastructure will be used by European startups, SMEs, or corporates.
We must be responsible when investing taxpayer money in a new technology field at the proposed scale. The potential of AI to enhance productivity and competitiveness may be vast, but it remains largely unrealized and unproven. The US, despite leading the world in AI deployment, is yet to see real GDP impact, as evidenced by the most recent update showing a 0.3% contraction in Q1-2025. Given that AI has been deployed since 2022, one would expect a larger impact on the GDP development.
For Poland specifically, championing this debt-instrument approach offers several strategic advantages: it positions Poland as a thought leader in responsible AI investment within the EU; it creates a model that could prioritize infrastructure development in Central and Eastern European member states; and it establishes a financing framework that could better support Polish AI startups and SMEs through guaranteed infrastructure access. By using a debt-instrument, we can let the market determine the risks while providing the additional financial capacity needed to stimulate AI infrastructure development that benefits all European citizens, including those in emerging tech ecosystems like Poland's.