.png){kind=logo}
AI leaders are raising concerns about the technology’s potential risks to humanity, even as the US and China intensify their competition. Behind the warnings is a wider debate over regulation, investment, and whether America’s costly AI strategy can withstand cheaper Chinese models.
The latest debate began on September 8 after former Anthropic researcher Jacob Coxon wrote on X that people developing AI genuinely believe the technology could destroy humanity by the end of the decade. His post was viewed 173 million times, bringing wider attention to concerns about AI development and safety.
On September 12, Anthropic CEO Dario Amodei published an essay calling for greater restraint in developing AI capabilities. OpenAI CEO Sam Altman and SpaceX founder Elon Musk also called for slower development or stronger government oversight.
The warnings have emerged as the US and China enter another phase of their AI competition. US technology companies are investing heavily in AI infrastructure, while Chinese firms are developing lower-cost models that are increasingly challenging American products.
Financial concerns are adding to the debate. Morgan Stanley estimates that US hyperscalers, along with NVIDIA and Broadcom, have more than USD 3 trillion in off-balance-sheet liabilities and guarantees.
US hyperscalers are expected to invest as much as USD 1.5 trillion in artificial intelligence next year. Oxford Economics' Ryan Sweet said building AI is expensive but could still prove worthwhile.
However, Apollo Global Management's Torsten Slok estimates that hyperscalers' operating cash flow would need to rise from USD 600 billion last year to USD 2 trillion in 2030 to support those investments.
The economics of AI are becoming another concern for US companies. The report said one of Chinese startup DeepSeek's flagship models may cost more than 100 times less to operate than Anthropic's Claude Fable 5. Products from other Chinese companies can cost up to 50 times less than leading American alternatives.
At the same time, the performance gap between Chinese models and the most expensive American systems is narrowing. This could make it harder for US companies to justify their rapidly rising infrastructure spending.
Also Read: China Probes DeepSeek, Moonshot AI Over Claude Data Routing Claims
Investors remain divided over whether the enormous investments made by American technology companies will generate sufficient returns. The Nasdaq's recovery to a recent record high indicates that confidence remains strong, although skeptics continue to question the economics.
Investor Steve Eisman has argued that AI companies could use concerns about the technology's risks to encourage regulation that creates barriers for competitors.
Meanwhile, US Treasury Secretary Scott Bessent said Washington and Beijing had agreed to establish an official dialogue on artificial intelligence, with senior officials expected to meet in China by the end of the year.
The debate now extends beyond AI safety. It also concerns whether the US can maintain its technological lead while competing against Chinese companies offering increasingly capable and cheaper AI systems.
