Stonepeak CEO flags capital markets as biggest risk to AI infrastructure investment boom
The rapid expansion of AI infrastructure continues to attract strong backing from banks, private equity firms and capital markets, but Stonepeak Infrastructure Partners CEO Michael Dorrell has warned that sustained access to financing remains the sector's greatest long-term vulnerability, according to a report by Bloomberg.
Stonepeak Infrastructure Partners CEO Michael Dorrell has cautioned that access to financing remains the sector's biggest long-term risk to the rapid expansion of AI infrastructure, according to Bloomberg. Despite strong support from banks, private equity firms and capital markets, Dorrell suggested that investment in AI-related infrastructure projects may not persist indefinitely due to the lack of evidence that investors are pulling back, even with unprecedented capital flowing into the sector.
He stressed the critical dependence of the AI infrastructure boom on open capital markets, warning that if financing conditions tighten significantly, businesses reliant on external funding could struggle to complete projects or execute growth strategies. The consultancy McKinsey estimates that global spending on data centre infrastructure could reach a staggering $5.2tn by 2030, fueled by the growing demand for AI computing capacity.
Dorrell differentiated today's investment cycle from the technology bubble of the late 1990s, emphasizing that many modern data centre developments are backed by long-term customer contracts rather than speculative demand, providing a stronger commercial foundation. He also noted that increasing political and regulatory scrutiny of new data centre developments in the US, such as restrictions on hyperscale data centre construction in states like New York, could pose challenges and create tensions between local planning concerns and the US government's strategic goal of maintaining AI leadership over China.
Written by urgent.news from Private Equity Wire's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
