Urgent.News

What's breaking now, across thousands of outlets.

AI

An AI ‘debt bomb’ crisis? No. This isn’t Enron 2.0 | Gene Marks

Fears of a datacenter buildout debt crisis are exaggerated. The risks are different than in the past and they are recoverable Some experts are warning of a looming “debt bomb” crisis because big datacenter builders such as Meta , Oracle , xAI and CoreWeave are not only raising billions to construct these facilities but are also not recognizing these long-term debt obligations on their balance…

An AI ‘debt bomb’ crisis? No. This isn’t Enron 2.0 | Gene Marks

Concerns about a looming "debt bomb" crisis due to big tech companies like Meta, Oracle, xAI, and CoreWeave avoiding long-term debt obligations on their balance sheets have been exaggerated, according to financial experts. These companies are constructing datacenters to support their growing AI and cloud computing needs and financing these projects through separate entities known as special-purpose vehicles.

This allows the tech giants to claim ownership of the datacenters while keeping the debt off their primary financial statements.

The Financial Times reported in December 2025 that tech companies have shifted over $120 billion of AI datacenter spending off their balance sheets using these financing structures. Goldman Sachs estimates that hyperscalers could spend $5.3 trillion on AI and datacenters by 2030, with private markets playing an increasingly important role in financing these projects.

Some experts are worried that these companies are not properly disclosing the long-term impact of their debt, drawing parallels to the Enron scandal in 2001. However, the risks associated with these financing structures today are different. The disclosure requirements are more stringent, the scrutiny is intense, and the investing public is more informed.

Additionally, unlike the biotech sector in the 1980s and 1990s, where many drugs failed in clinical testing, datacenters provide a legitimate market need. AI adoption is still in its early stages, with only 17.8% of the working-age population using generative AI, according to Microsoft estimates.

While some investments may fail, lenders may lose money, and some datacenters might be worth less than their original purchase price, the rationale behind these financing structures is to spread capital requirements and risk among investors willing to take them. The assets are real, and demand for computing capacity remains strong. Therefore, the author concludes that there is no reason to be concerned about a "debt bomb" crisis in the tech industry.

Written by urgent.news from Guardian Business's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Read the original at theguardian.com →

More in AI

How a Dedup Pass Deleted My Training Curriculum

Code: Megapixel99/capture-the-flag I built a capture-the-flag arena where language models attack and defend each other's containers.

  • Qwen2.5-3B-Instruct trained on capture flags
  • Defense performance struggled, attempts rolled back
  • Deduplication stage discarded weights, leaving binary filter

More from Sunday 23 August →