Banks to offload $15bn debt for US AI data centre, raising questions for UK financing models
Executive summary
US banks are planning to offload $15bn of debt associated with a Google-backed Anthropic data centre in Texas. This move indicates increasing caution among Wall Street lenders regarding the substantial financing required for AI infrastructure, with a shift towards the bond market for such large-scale projects.
Reporting based on ft.com
Why it matters
While the project is in the US, the immense capital requirements and evolving financing strategies for AI data centres, as highlighted by this situation, offer crucial insights for the UK. It suggests potential challenges and innovative funding models that may emerge as the UK accelerates its own AI infrastructure development, impacting everything from investment to grid stability.
Sector impact
Analysis by AI Energy Intelligence UK
The article reports that the new Texas campus will be powered by its own natural gas power plant to avoid delays and surging costs, acknowledging rising concerns around strained power supplies and utility prices in the region due to numerous data centre projects. This highlights the significant energy demands of large AI data centres and the potential for these facilities to influence local energy generation strategies and grid stress.
The project's reliance on a dedicated natural gas power plant for its 2,000-acre campus demonstrates a strategy to ensure power supply security and avoid grid delays. This approach underscores the critical importance of reliable power for AI infrastructure and suggests a potential trend towards integrated energy solutions to mitigate risks associated with grid strain.
Banks are seeking to offload $15bn in debt for this data centre, with broader efforts to find buyers for over $50bn of construction debt for other data centre projects. This indicates that the scale of AI infrastructure financing is stretching traditional banking limits, pushing lenders towards the bond market and potentially leading to higher yields for investors willing to take on construction and cost overrun risks.
Key statistics
Figures as reported by ft.com. See original source for context.
Long-term implications
The shift from bank loans to the bond market for financing mega AI projects could establish a new precedent for how such infrastructure is funded globally. This could lead to a broader market for AI infrastructure bonds, offering alternative investment opportunities but also requiring investors to assume more direct construction and operational risks.
Frequently asked questions
What is the key financial development discussed?
Banks are planning to offload $15bn of debt related to a Google-backed data centre in Texas that will be leased to Anthropic, moving towards the bond market for refinancing.
Why are banks offloading this debt?
Wall Street lenders are becoming wary of holding AI infrastructure debt on their balance sheets, and offloading it helps reduce their AI risk exposure and frees up capacity for more lending.
What are the power implications of this data centre?
The new Texas data centre campus will be powered by its own natural gas power plant to mitigate delays and surging costs, addressing concerns about strained power and water supplies.
Explore related tools
Original source
This story summarises reporting from ft.com. Read the original for full context.
Read on ft.comRecommended reports
Independent UK research that expands on the themes in this story.
The definitive UK view of AI's electricity, grid and infrastructure impact in 2026. An independent, evidence-based report on demand growth, data centre build-out, AI Growth Zones and the trajectory to 2035 — for UK policymakers, operators and investors.
Read the reportThe environmental cost of AI-generated answers versus conventional search, benchmarked for UK decision-makers. An independent, evidence-based comparison of electricity, water and carbon per query across Google, Gemini, ChatGPT, Copilot, Perplexity and Claude — with forecasts to 2035.
Read the reportBottom-up model of AI-driven electricity growth across UK regions to 2030, with sensitivity to model efficiency and adoption.
Read the reportGet the UK AI Energy briefing
Analysis on AI, electricity and the UK grid, straight to your inbox.