Executive Summary
Greg Jensen, co-Chief Investment Officer of Bridgewater Associates, has issued a stark warning that the real artificial intelligence bubble has not yet formed, asserting that investors are underestimating the impending scale of the AI transformation. His commentary comes as economic data reveals that AI-related capital expenditure (capex) is responsible for over half of U.S. GDP growth in the first half of 2025. This intense, concentrated investment has propped up the economy but has also created a fragile environment, with market gains narrowly focused on a few infrastructure providers and raising concerns about systemic risk should this spending contract.
The Event in Detail
According to Jensen, the market has not yet entered the speculative frenzy typical of a major technological shift. Instead, he argues that the world is now entering a "more dangerous phase" of the AI cycle. This next stage is defined by intensifying competition, accelerating capital spending, and an ensuing grab for scarce resources, for which he believes investors are unprepared. This view challenges the narrative that the current AI rally is a simple repeat of the dot-com bubble, suggesting the most volatile period is still ahead.
Market Implications
The economic reliance on AI investment is profound. According to multiple economic analyses, the AI boom accounted for more than half of the U.S. economy's 1.6% growth rate in the first six months of 2025. Stephen Juneau, an economist at Bank of America, noted that AI is "the only source of investment right now," as private business investment excluding AI has been stagnant since 2019. This concentration creates significant economic vulnerability. Jonathan Millar, Senior U.S. Economist at Barclays, estimates that a 20% to 30% decline in the stock market could reduce annual GDP growth by one to 1.5 percentage points, highlighting the economy's growing dependency on the performance of AI-linked equities.
Bridgewater's internal research supports Jensen's caution, noting that the stock rally has narrowly benefited companies at the center of the AI infrastructure buildout, such as Nvidia. The firm's analysis indicates that price increases in key stocks like Nvidia have been largely justified by announced capex plans and corresponding near-term earnings expectations. However, it also points out that very little potential upside is priced in for the companies that will benefit from the broader deployment of AI across the economy. This view is echoed by other market strategists who suggest looking toward sectors like industrials and utilities, which are part of the AI revolution but do not carry the same stretched valuations as semiconductor firms. Major technology firms, including Microsoft, Alphabet (Google), Meta Platforms, and Amazon, have announced massive capex plans for data centers and AI hardware, fueling this infrastructure-centric boom.
Broader Context
The current economic landscape is characterized by a growth cycle heavily reliant on a single theme: AI infrastructure spending. While this has driven positive GDP numbers and corporate earnings for a select group of companies, it has also created a precarious economic dependency. The "dangerous phase" articulated by Jensen refers to the potential fallout if these massive capex commitments slow down or if market sentiment shifts. The concentration of gains in a handful of high-valuation technology stocks creates a fragile system where a correction could have outsized macroeconomic consequences, turning a sector-specific downturn into a broader economic drag.
