Meta rating outlook cut fuels AI investment risk, market funding alarm

Astronomical investments poured into the artificial intelligence (AI) supremacy race are draining big tech companies' free cash flow and surfacing as credit risk. As near-unlimited liquidity dries up, companies with weak cash generation face shrinking funding limits and are forced to rely on high-interest subordinated loans.

The first warning light has flashed at Meta Platforms. S&P Global Ratings on Oct. 30 abruptly cut its rating outlook on Meta Platforms to negative from stable. The credit rating itself was kept at AA-, but the assessment is that explosive AI infrastructure spending and the resulting sharp deterioration in free cash flow (FCF) are undermining financial soundness. Meta Platforms plans to invest up to $145 billion in AI facilities and other areas in 2026. But its free cash flow swung to a deficit of $2.12 billion in the third quarter.

Credit rating agencies see this not as a one-off expense but as a structural distortion of long-term cash flow. S&P Global Ratings projected Meta Platforms' free cash flow deficit will balloon to $71 billion in 2027 and still reach $39 billion in 2028. That reverses earlier forecasts that it could turn positive in 2028. Unlike Microsoft, Google and Amazon, which recoup cash by leasing out data center and cloud capacity to external customers, Meta Platforms relies entirely on its own services and online advertising revenue. This extremely limited ability to recover investment costs was the key reason for the outlook downgrade.

Meta Platforms pushed back, saying it disagrees with the rating agency's analysis of its AI monetization capabilities and that business opportunities remain to support strong creditworthiness. But the fact that it is pursuing a record $95 billion share issuance to fund capital spending shows it cannot keep up with the pace of investment using internally generated cash alone. Market doubts persist that financial pressure has grown so much that it must raise capital even at the cost of massive equity dilution.

This funding squeeze is spreading to the core AI developers at the top of the startup ecosystem. According to the Financial Times and others on Oct. 30, OpenAI slashed its new fundraising target to $20 billion from an initial $30 billion. Its expected valuation was also cut by $150 billion to $1.25 trillion from $1.4 trillion. The amount secured at the first commitment deadline was only $11 billion, less than half the original target, and the company is now scrambling to fill the remainder by bringing in family offices and high-net-worth individuals.

As the credit rating system begins to function fully, the bond market's bar is rising sharply. For Anthropic, which is preparing for an initial public offering (IPO), investment banks and private equity firms have formally warned that its post-listing credit rating is likely to fall short of investment grade (BBB- or higher) and remain at speculative-grade BB. With Anthropic having announced more than $800 billion in infrastructure investment through 2045, a speculative-grade label would completely block bond purchases by conservative institutional investors such as pension funds and insurers. That could lead to astronomical borrowing costs and deal a fatal blow to business sustainability.

As funding dries up, even distorted forms of credit enhancement are emerging. In the $63 billion financial structure designed for Anthropic to procure Google AI chips, Broadcom, a semiconductor supplier with a strong credit rating, guaranteed the entire senior loan of more than $42 billion. Half of the $18 billion high-risk subordinated loan without guarantees was taken by private equity firm Blackstone at an annual yield above 10%. As financial institutions do not trust AI developers' repayment capacity, a semiconductor seller is taking on credit risk to boost sales while private equity funds pocket high-interest premiums — an unconventional circular financing structure has become reality.

The premium once backed by aggressive investment is gone, and the cold standards of the bond market — cash erosion and debt burden — are now being applied. If capital spending continues to outpace profit generation, a downgrade of Meta Platforms' actual credit rating and a chain of cracks in financing structures propped up by partner guarantees cannot be ruled out. The interest rate on Anthropic's follow-on capital raising through year-end and whether Meta Platforms narrows its free cash flow deficit in the first half of next year are set to be key indicators determining a credit crunch across the tech sector.

※ This article is an AI translation of the Korean original published by Korea Credit News; it was checked by automated quality guards and published by the English edition desk.