The leaders of major AI companies are calling for a more cautious pace of development, as soaring investment, mounting financial exposure and intensifying competition from China raise questions about the sustainability of the industry’s current model.
Executives at some of the world’s biggest artificial intelligence companies have recently begun calling for greater caution in the development of increasingly powerful AI systems.
The public justification is familiar: concerns that AI could become difficult to control and create risks ranging from attacks on financial systems to the development of dangerous weapons or the spread of false information capable of triggering geopolitical crises.
But another concern is emerging alongside the debate over AI safety — the enormous financial exposure created by the industry’s race to build increasingly powerful models and infrastructure.
A $3.1 trillion financial exposure
OpenAI, Anthropic, Meta and Google are among the companies investing hundreds of billions of dollars in AI infrastructure, including data centres and advanced chips.
According to estimates cited in the analysis, the seven leading AI developers and chipmakers have accumulated around $3.1 trillion in off-balance-sheet debt and customer financing arrangements. The figure is approaching the scale of the property-related debt that was at the centre of the 2007-08 financial crisis, although the AI exposure has accumulated over a much shorter period.
Nvidia, the world’s most valuable company by market capitalisation, has also become deeply involved in financing arrangements linked to AI infrastructure. According to the Economist, its commitments related to financing, leasing and potential repurchases of its chips could represent at least $300 billion in off-balance-sheet exposure.
The enormous cost of training and running AI models is forcing companies to make long-term commitments to computing capacity, in some cases involving data centres that have yet to be built.
OpenAI is also reported to be losing around $10 billion a month as it funds its expansion, while Anthropic is preparing for a potential stock-market listing.
China changes the equation
The financial pressure is being compounded by China’s rapid progress in artificial intelligence.
Chinese companies have developed increasingly capable, lower-cost and often open-source models, challenging the business model of US companies that have invested heavily in proprietary AI systems.
Models such as DeepSeek and Kimi have demonstrated that advanced AI capabilities can be offered at a fraction of the cost associated with some US-developed systems. The growing availability of open-source alternatives could make it more difficult for American AI companies to justify the enormous infrastructure investments currently being made.
The competitive threat is not limited to software. China is also seeking to build an alternative ecosystem for AI hardware, potentially challenging Nvidia’s dominance in advanced chips.
Palantir’s European CEO Louis Mosley has argued that China’s strategy is aimed not only at winning the technological race but also at undermining the economic model of leading US AI companies.
Wall Street exposure
The implications could extend well beyond the technology sector.
AI-related companies have been a major driver of the US stock market’s gains in recent years, leaving investors increasingly exposed to the sector’s valuations. A sharp decline in AI-related stocks could therefore have consequences for the wider financial system, particularly if it coincides with the large debt and infrastructure commitments accumulated by the industry.
A 20% or 30% collapse in AI valuations would potentially erase trillions of dollars in market value and could weaken corporate investment and consumer spending if the decline became sufficiently broad.
The comparison with the aftermath of the dot-com bubble has consequently returned to the debate, although the scale and structure of the current AI investment cycle are different.
Why are AI leaders calling for caution?
AI executives including Anthropic’s Dario Amodei and OpenAI’s Sam Altman have long warned about the potential risks associated with increasingly powerful AI systems. At the same time, they have argued that slowing down unilaterally would be difficult while geopolitical rivals, particularly China, continue to develop the technology rapidly.
The recent calls for a more measured approach therefore come at a moment when the industry’s economic model is facing growing scrutiny.
The central question is no longer simply how quickly AI can be developed. It is also whether the companies leading the race can sustain the enormous capital expenditure required to remain at the frontier while facing increasingly powerful and cheaper competitors.
For some observers, the renewed emphasis on AI risks reflects genuine concerns about the technology’s potential consequences. Others argue that the industry’s financial pressures are an equally important part of the story.
Either way, the AI race is increasingly becoming a contest not only over technological capabilities, but also over capital, infrastructure, business models and the ability to withstand financial pressure.
Be First to Comment