AI Debt Crisis Could Send Bitcoin to $1M, Predicts Arthur Hayes
Arthur Hayes, chief investment officer at Maelstrom, has reaffirmed his ambitious prediction that Bitcoin could reach $1 million by 2030. According to Hayes, a debt-fueled artificial intelligence boom is set to collide with its own vulnerabilities in late 2027 or early 2028, potentially triggering a chain of events that benefits Bitcoin.
Hayes' argument centers on the enormous amounts of credit being used to finance AI-related projects such as data centers, chips, and computing infrastructure. He warns that if these projects struggle to generate enough cash to service their debts, losses could spread to banks, private-credit funds, and other lenders.
Governments and central banks may respond by injecting liquidity into the financial system, which Hayes believes would ultimately benefit scarce assets like Bitcoin. The IMF has already highlighted concerns about data-center capital expenditure, with Morgan Stanley estimates suggesting it could reach $2.9 trillion through 2028 - far exceeding available cash flow from hyperscalers.
The financing side of the AI boom is attracting increasing scrutiny, with investors becoming more selective about lending to riskier AI businesses. As of September 30, AI-related borrowing in the US leveraged-finance market had reached approximately $88 billion in 2026, up from around $20 billion at the start of 2025.