FT podcast links rising US debt to higher AI borrowing costs

In short: A Financial Times podcast argues that rising US government debt could raise interest rates and make it harder for AI companies to fund their expansion.

What's going on

In a new episode of The Economics Show with Soumaya Keynes, investor and commentator Ruchir Sharma says two big borrowing sprees may be heading for a clash. Tech companies are borrowing large amounts to build AI systems, and the US government is borrowing large amounts to pay for its spending.

A key part of the AI boom is building data centres, which are warehouses full of computers that run AI services. These are expensive to build and run, so companies often rely on loans and other financing.

Sharma says investors are already nervous about two things at the same time. One is the size of US government debt. The other is the fear that excitement about AI has pushed prices and expectations too high.

He points to US Treasury yields as an important link between the two issues. A Treasury yield is basically the interest rate the US government has to pay to borrow money. If those yields rise, other borrowing often gets more expensive too, like a benchmark that pulls other rates upward.

What to watch

If Treasury yields keep rising, AI companies may face higher costs to borrow money for data centres and related projects. That could slow down spending on AI infrastructure, even if demand for AI tools stays strong. Sharma says investors should keep a close eye on Treasury yields as a simple number that can signal whether financing conditions are getting tighter.

Source: Financial Times