JP Morgan CEO says hyperscaler AI spending could hit $1T in 2027

JPMorgan CEO Jamie Dimon said hyperscaler spending on artificial intelligence could reach $1 trillion in 2027, up from roughly $700 billion this year and about $300 billion last year. He argued that the surge - driven by hiring, new data centers, power plants and equipment purchases - is already boosting growth and could add roughly 1 percentage point to GDP annually while exerting some upward pressure on inflation in the near term. Over the long run he expects AI to be deflationary because of productivity gains, but warned it’s too early to pick clear winners, likening the landscape to the internet bubble where unexpected companies emerged dominant. He also noted that many AI investments are “table stakes” and that benefits such as improved customer experience are hard to quantify.

Dimon flagged broader macro and geopolitical implications: heavy capital demand from infrastructure, rearmament and persistent government deficits could push interest rates higher, and inflation might not fall quickly - the Fed should stick to a 2% target. He urged full U.S.-China engagement on trade, AI and security ahead of the leaders’ meeting, called for renewed U.S.-India trade talks and predicted India’s economy could triple over the next decade, while urging nuanced U.S. policy on India’s purchases of Russian oil to avoid harming global markets.