Macro
Carlyle Group's Jason Thomas Warns AI Infrastructure Boom Mirrors Pre-Crisis Mortgage Trends
Jason Thomas, Managing Director and Head of Global Research and Investment Strategy at Carlyle Group, has raised alarms regarding the current AI infrastructure lending boom, likening it to the mortgage finance strategies that preceded the 2008 financial crisis. In his report titled “Let Them Eat Compute: Data Centers’ Implications for Mortgage Finance,” Thomas argues that the rapid influx of capital into AI-related infrastructure is distorting credit markets, leading to higher interest rates and sidelining traditional borrowers, particularly in the housing sector.
The scale of AI infrastructure spending is staggering, with estimates suggesting hundreds of billions of dollars are being invested quarterly, growing at an annualized rate of 40-60%. According to Carlyle’s data, hyperscale companies have increased their property, plant, and equipment by 50-200% since late 2023. This surge in investment comes at a time when the US 30-year fixed mortgage rate stood at 6.58% as of August 2025, creating a challenging environment for new homebuyers.
Thomas emphasizes that existing homeowners with lower locked-in rates are benefiting, leading to a two-tier housing market. He also points out that federal deficits are currently 1.5 times larger than the average from 2010 to 2019, further intensifying competition for available capital.
The crux of Thomas's warning lies in lender psychology. Today's data center financing relies on the assumption that AI demand will continue to rise, and that the compute capacity being developed will attract paying customers. However, this reliance on future demand raises concerns about the sustainability of revenue streams and the potential for mispricing risks. A few hyperscale companies dominate this buildout, which could amplify the consequences of any downturn.
While Carlyle has seen success in this sector, reporting a fivefold return from the 2026 sale of its Copia Power platform, Thomas's report serves as a cautionary note. The ongoing transformation of capital markets driven by AI infrastructure demand may not be accurately reflecting the associated risks.
FAQ
What warning did Jason Thomas from Carlyle Group issue regarding AI infrastructure lending?
Jason Thomas warned that the current AI infrastructure lending boom resembles the mortgage finance strategies that led to the 2008 financial crisis, suggesting that it is distorting credit markets and could lead to negative consequences.
How much capital is being invested in AI infrastructure according to the report?
Estimates suggest that hundreds of billions of dollars are being invested quarterly in AI-related infrastructure, with growth rates of 40-60% annually.
What impact is the AI infrastructure spending having on the housing market?
The influx of capital into AI infrastructure is contributing to higher interest rates, which is sidelining traditional borrowers in the housing sector and creating a two-tier housing market.
What concerns does Thomas raise about the sustainability of AI infrastructure investments?
Thomas expresses concerns that the reliance on future AI demand for data center financing may lead to mispricing risks and unsustainable revenue streams, especially given that a few hyperscale companies dominate the market.
What does Thomas's report suggest about the current state of federal deficits?
Thomas notes that current federal deficits are 1.5 times larger than the average from 2010 to 2019, which intensifies competition for available capital and complicates the financial landscape.