A debate is intensifying over why leading artificial intelligence executives have called one after another for slowing the development of advanced AI. Critics say the real motive lies in the interests of companies preparing for initial public offerings, which stand to reduce financial burdens and block latecomers from catching up.
The Wall Street Journal reported on the 14th that the clash between money and safety has created an enormous crisis for AI, judging from the situations facing major AI companies such as Anthropic and OpenAI. The newspaper said the entire industry is at risk as scientific progress, moral obligation and economic incentives collide.
The debate was ignited by Anthropic Chief Executive Dario Amodei. He recently said AI is approaching a stage where it can improve its own performance, and urged that the pace of development be adjusted so that safeguards can keep up with technological progress. OpenAI CEO Sam Altman, xAI CEO Elon Musk and Google DeepMind CEO Demis Hassabis echoed the call. Microsoft, meanwhile, recently unveiled an AI code of conduct aimed at keeping AI from threatening humanity.
But some in the U.S. government and the investment community argue that the push for AI brakes cannot be seen purely as a safety argument. David Sacks, the White House science and technology adviser, said that if leading AI companies were genuinely worried about the risks, they could simply slow development on their own. Demanding that the government impose the regulations they want, he said, can look like regulatory capture.
Some analysts say a slower pace would also help the financial structure of the front-runners. Developing cutting-edge models requires enormous spending on graphics processing units and data centers, but easing the development race would cut those costs while preserving revenue from existing models. For Anthropic and OpenAI in particular, both of which are pursuing listings, it could buy time to improve profitability.
Despite the recent controversy, Anthropic is accelerating its push for a Nasdaq listing in late October this year. It has secured Nvidia as a key investor and is in talks to raise $10 billion. Altman has said a listing this year would not be appropriate for OpenAI because of safety-related work, but critics say the real purpose is to avoid a head-to-head fight with Anthropic. OpenAI has raised its projection for spending on AI computing infrastructure through 2030 to $750 billion from $600 billion. Microsoft's pre-emptive announcement of an ethics code is also read as a strategic move by a company that has fallen behind in the technology race to get ahead in the battle over standards.
The U.S.-China contest for AI supremacy is another factor that undercuts how realistic a slowdown would be, given concerns that Chinese players such as DeepSeek and Moonshot AI could narrow the technology gap while U.S. companies hold back. President Donald Trump has also taken a negative view of the industry's slowdown argument, saying AI regulation could benefit only China.
