Investors are growing more anxious about the scale of artificial intelligence spending, with large borrowing plans and infrastructure commitments drawing scrutiny as interest rates rise. Demand for financial protection against a reversal is also increasing, reflecting concern about the risks attached to the AI investment boom.
Alphabet, Meta and Oracle have either issued new bonds or reportedly prepared to do so. The borrowing activity comes as investors examine whether the industry’s expanding infrastructure commitments can generate enough revenue to justify their cost. Higher interest rates add to that scrutiny by making new debt more expensive and raising the returns investors can earn from competing investments.
Valuation concerns extend beyond publicly traded technology companies. The Wall Street Journal reported that institutional investors have urged Anthropic to reduce its valuation target for an initial public offering. The report described the AI developer as expecting to spend nearly $900 billion on infrastructure over the coming years. That figure represents an expected spending commitment, rather than money already deployed.
Investors have also increased their demand for options and credit-default swaps that can provide protection against adverse market moves. Options can be structured to offset losses when share prices decline, while credit-default swaps transfer specified credit risks associated with a borrower. Growing interest in those instruments points to a desire to limit exposure, although hedging alone does not establish that investors expect an imminent collapse.
PIMCO’s chief executive warned that a future downturn could expose vulnerabilities built up during the expansion. The concern is not simply whether AI technology will prove useful, but whether the pace of investment, the prices paid for businesses and the financing used to support construction will withstand weaker conditions.
AI infrastructure requires substantial spending before the resulting capacity can earn revenue. Training and operating advanced models depend on specialized processors, networking equipment, data centers and electricity. Those requirements can create a gap between the timing of cash outlays and the income needed to support them. Borrowing can help bridge that gap, but it also creates interest and repayment obligations regardless of how quickly customer demand develops.
The economics differ across the businesses involved. Established technology groups can finance AI expansion alongside existing operations, while an AI developer seeking a public listing faces a separate test of how public-market investors value its growth prospects and funding needs. An IPO valuation target is a negotiating position, not a completed market valuation.
For investors, the unresolved issue is how closely infrastructure spending will track commercially sustainable demand. The reported pressure on Anthropic’s IPO ambitions and the increased use of protective trades show that financing terms and downside exposure are receiving greater attention alongside the industry’s expansion plans.
Sources: Semafor