Developing Story
Carlyle's Jason Thomas – AI Financing 'Echoes' Pre-Crisis Mortgage Lending
Carlyle's Jason Thomas warned that AI data center financing structures resemble pre-2008 mortgage lending practices and said an AI investment slowdown is 'more likely than not,' adding a prominent voice to growing Wall Street skepticism about AI infrastructure financing sustainability.
This page was written and is updated by the paper from the source articles listed below. Importance is how central the topic is to current coverage; confidence is how well-sourced the page is; mentions counts the articles that fed it. How the wiki works →
## Overview
Jason Thomas, head of global research and investment strategy at Carlyle Group, warned that lenders' approach to financing AI data centers has a 'troubling similarity' to the mortgage lending model that preceded the 2008 financial crisis (Bloomberg, September 15). Speaking from Carlyle's Global Investor Conference in Washington, Thomas separately said an AI slowdown is 'more likely than not' (Bloomberg, September 15).
## Key Claims
- Data center financing structures reportedly resemble pre-crisis mortgage securitization patterns, according to Thomas (Bloomberg, September 15).
- Thomas said the Federal Reserve is under 'enormous pressure' to deliver a 25 basis-point rate cut/hike (as reported) (Bloomberg, September 15).
- He characterized an AI investment slowdown as more likely than not, amid a broader debate among investors about the pace of the AI buildout (Bloomberg, September 15).
## Context
This warning arrives amid a broader wave of commentary from investors and analysts (including George Noble of Noble Capital Advisors, who separately called AI 'the biggest reallocation of capital in history' while flagging bear-case risks) about the sustainability of AI infrastructure financing, following a string of large-scale AI data center debt issuances and buildouts by hyperscalers. The mortgage-crisis analogy specifically targets underwriting practices for data center loans and their resemblance to structured mortgage products that failed in 2008.
## Significance
This represents an important data point in the growing 'AI bubble' debate among institutional investors, sitting alongside other 2026 developments such as record AI infrastructure spending, junk-bond financing of AI infrastructure (Core Scientific), and significant risk transfer hedges by banks (TD Bank). Should this narrative gain traction, it could affect credit spreads on data center debt, bank risk management practices, and regulatory scrutiny of AI infrastructure financing structures.
## What to Watch
- Whether other major institutional investors echo the mortgage-crisis comparison
- Regulatory response (Fed, bank supervisors) to data center lending practices
- Performance of AI-related structured credit products
- Broader market reaction to AI slowdown warnings