Organizations are expending massive amounts of capital to stay competitive in the AI race. Debt levels are increasing. In the midst of this activity, Amazon has entered an agreement to secure approximately $17.5 billion from several financial institutions, as reported by Bloomberg.

The financial institutions involved in the loan reportedly include Citigroup, JPMorgan Chase, Wells Fargo, HSBC, and BofA Securities. This arrangement has been identified as a delayed draw term loan, which allows Amazon to access the funds at its discretion instead of receiving the entire amount at once, providing flexibility in how and when the funds are utilized.

This loan follows shortly after the announcement that Amazon is set to raise $14 billion via a Canadian bond issuance, bringing its total new financing to around $31.5 billion within approximately 48 hours.

It remains unclear how Amazon intends to allocate this influx of capital. Reuters mentions that the new loan will cater to “general corporate purposes.” TechCrunch has reached out to Amazon for additional details.

Amazon is not alone in this endeavor. To finance new AI infrastructure such as chips and data centers, companies are utilizing unprecedented capital expenditures. More companies are borrowing funds to support their extensive AI expansions. The pressing question for investors and analysts is not whether this expenditure is essential — but rather if the returns will ever validate it.

The magnitude of the borrowing is remarkable even by Silicon Valley benchmarks. About a week prior, Alphabet, Google’s parent company, announced plans to generate $80 billion through a stock sale aimed at “funding its investments in a balanced manner while maintaining a solid balance sheet.” Meta has also revealed plans to secure $30 billion through a bond sale — its largest amount to date.