AI Development Slowdown Could Create Bitcoin-Friendly Liquidity, Says Arthur Hayes

Arthur Hayes, co-founder of BitMEX and a prominent Bitcoin optimist, has stated that the recent slowdown in AI development by major companies may be driven more by profitability concerns than safety issues, potentially creating a favorable liquidity environment for Bitcoin if financial markets face shocks from AI-related debt.
According to a report by CoinPost on October 22 (local time), Hayes discussed in his newsletter 'Crypto Trader Digest' that companies like Anthropic, OpenAI, and SpaceX are adjusting the pace of their general artificial intelligence (AGI) development. While these firms cite safety as the reason for slowing down, Hayes argues it reflects the high costs and deteriorating profitability of AI development.
He noted that U.S. AI companies' computational demand is supporting over $1 trillion in investment-grade debt alone, with additional billions tied to lower-rated corporate bonds and loans. If AI development slows, this could impact not only technology timelines but also the value of financial assets linked to AI data centers.
Hayes highlighted that U.S. AI firms are facing profit pressure from competition with cheaper Chinese AI models. As demand grows for affordable AI solutions beyond high-performance systems, he suggests these companies may shift strategies toward optimizing existing computing resources rather than investing heavily in new large-scale models.
This could reduce the need for computational capacity, affecting investment and financing in AI data centers. However, Hayes emphasized that much of this infrastructure is financed through debt, meaning a slowdown in credit demand could lead to declining asset values, particularly impacting insurance companies as vulnerable links in the chain.
Hayes cited analysis of mispriced assets, noting that private equity firms acquiring insurers and establishing captive reinsurers—often registered in lightly regulated states like Vermont—are using this structure to secure regulatory capital buffers. He warned that up to $154 billion in assets tied to AI data center debt or private credit linked to AI companies could be at risk, with limited external visibility into their actual health.
He referenced a case where a Brookfield-affiliated reinsurer reported assets valued at $148 million but disclosed no payment obligations to regulators. Hayes also warned that if AI computational demand slows faster than expected, credit rating agencies might downgrade related debt, forcing insurers to secure additional capital.
He suggested the U.S. government could intervene, either as a last buyer of AI computing resources for national security reasons or by supporting the insurance sector, similar to how it aided AIG during the 2008 financial crisis. Hayes believes such interventions could increase liquidity, potentially channeling funds into Bitcoin and other cryptocurrency markets.
However, Hayes stressed that this scenario remains speculative. The extent of AI investment slowdown, potential credit rating downgrades for AI-related debt, and how the U.S. government responds are still uncertain. His focus is on the financial market shocks and policy responses stemming from a potential AI development slowdown rather than the slowdown itself.
Korean Source
This article is an English localization of a Korean-language crypto news report. Original headline: AI 감속론에 숨은 금융 리스크?…아서 헤이즈 "비트코인엔 호재"