AI spending nears $1 trillion, but concerns grow over investment bubble

The conversation around artificial intelligence has shifted from enthusiastic promises to calculated concern, as global spending on AI approaches a staggering $1 trillion this year. This figure, representing about 1.8% of U.S. GDP, highlights AI's growing economic significance, according to Goldman Sachs. The focus is no longer solely on when AI will revolutionize everything, but rather on who will bear the cost if its economic benefits are slow to materialize.

“

The question being asked in financial circles is no longer 'When will AI change everything?', but rather: 'Who will bear the cost if its returns are delayed?'

Major financial institutions are questioning the structure of AI investments. A key concern is the rise of "circular financing," where chip manufacturers or data center operators invest in AI companies, which in turn commit to long-term contracts for chips or computing power from the same entities. The Bank for International Settlements has noted that these complex arrangements suffer from "limited disclosure," making risk assessment difficult for ordinary investors and regulators.

Credit rating agency Fitch has issued a stark warning, identifying the AI bubble alongside U.S.-China trade tensions as threats to global credit stability. This language from a rating agency signals a serious concern for lenders and pension fund managers. The market's anxiety is now extending beyond the scarcity of computing power to the potential for a surplus.

“

these complex structures... suffer from 'limited disclosure' which makes assessing real risks very difficult for the average investor and even regulatory bodies.

Warnings from major tech companies themselves highlight risks of overbuilding data centers and underutilization of capacity. Some heavily indebted clients are reportedly contracting for more computing power than they can realistically afford. This suggests that a portion of the demand driving AI's projected growth may be debt-financed rather than supported by sustainable revenue. Furthermore, the emergence of new competitors, particularly Chinese open-source models that approach U.S. performance at a lower development cost, challenges the assumption of U.S. technological dominance that justified high valuations for chip companies.

“

The AI bubble, along with U.S. trade tensions, now threatens global credit stability.