Some of the world’s biggest artificial intelligence companies are calling for a more cautious pace of development, raising questions over whether their concerns go beyond the potential risks of increasingly powerful AI systems.
Anthropic CEO Dario Amodei has called for a slower approach to frontier AI development, while OpenAI CEO Sam Altman has also backed the idea of greater caution. The companies have long warned about the possibility of AI being misused or becoming difficult to control, including scenarios involving cyberattacks, weapons development or critical financial infrastructure.
But critics and industry observers argue that another factor may be driving the recent calls for restraint: the enormous cost of the AI race and the growing challenge from China.
China is changing the competitive landscape
US technology companies have spent vast sums building increasingly powerful AI models and the data centres needed to run them. At the same time, Chinese developers have been producing increasingly capable models at significantly lower costs, with open-source systems offering an alternative to the closed models developed by companies such as OpenAI and Anthropic.
Models including DeepSeek and Kimi have highlighted the potential for Chinese companies to compete without matching the massive spending levels of their US rivals.
Palantir’s European CEO Louis Mosley has argued that China’s strategy is aimed in part at undermining the economic model of US frontier AI laboratories, describing the technological race as an economic competition as well.
The shift is particularly significant because the US AI industry has built its expansion around the assumption that increasingly powerful models would justify enormous investments in computing infrastructure.
Hundreds of billions committed to AI
OpenAI, Anthropic, Meta and Google are among the companies committing enormous amounts of capital to AI development and computing infrastructure.
Much of the spending involves long-term agreements for computing capacity, data centres and AI chips. Some of these arrangements involve complex financing structures and future commitments that may not be immediately visible on companies’ balance sheets.
Investment bank Morgan Stanley has estimated that off-balance-sheet financing and related exposure across major AI developers and chip companies could amount to trillions of dollars.
That has raised concerns about what could happen if AI revenues and valuations fail to keep pace with the industry’s investment boom.
Could the bigger risk be financial?
The rapid expansion of AI has also become an important driver of US stock-market valuations. A sharp correction in AI-related companies could therefore have consequences extending beyond the technology sector, particularly if falling valuations were accompanied by weaker investment and consumer spending.
The central question is whether the industry’s current spending model can be sustained while Chinese competitors continue to reduce the cost of AI development and expand their market presence.
Some industry figures continue to argue that the safety risks surrounding increasingly autonomous AI systems are genuine and require stronger safeguards. Others maintain that humans can retain effective control over AI agents.
The recent calls for a slower pace therefore reflect several overlapping concerns: AI safety, escalating infrastructure costs, intensifying competition from China and the financial risks created by the industry’s enormous investment cycle.
For companies that have spent hundreds of billions of dollars betting on the next phase of AI, the challenge may no longer be simply how quickly the technology can advance, but how long its current economic model can support that pace.
Be First to Comment