Trump Pushes Back Against AI Slowdown Calls, Warns of Losing to China

The clash over U.S. AI policy has come into sharp focus after Dario Amodei, CEO of Anthropic, published an open letter calling for slowing the pace of frontier development. Sam Altman of OpenAI and Elon Musk of xAI expressed support, while Demis Hassabis of Alphabet offered conditional backing. In contrast, Donald Trump and House Speaker Mike Johnson took the position that calls for a slowdown are an overreaction. According to reporting by Terrence O’Brien of The Verge, the two are wary of China’s rapid catch-up and oppose pausing development. This article is based on The Verge (All Rights Reserved) and relies on fair quotation under Article 32 of the Japanese Copyright Act.

Amodei’s Open Letter and Industry Support

Amodei published a lengthy open letter on September 12. Its central theme is pacing the frontier. The key phrase in the original is as follows.

pace the frontier

He expressed concern that capability gains in frontier models could outpace safety verification. His concern is that intensifying development competition could leave evaluation and safeguards behind. The open letter was issued as a call to the industry as a whole.

In response, Altman voiced support on X. Musk likewise expressed agreement. Both lead the development of large-scale models. Hassabis also lent tentative support to the proposal. Leaders of major research organizations thus appeared aligned.

This move is unusual. Because leaders of competing companies have aligned on slowing down. Behind it lies concern over the risks of high-performance models. There is shared vigilance about misuse and expanding autonomy. The open letter has triggered policy debate.

The race to expand AI infrastructure is also accompanied by improvements in operating environments. As shown in Microsoft Announces Solara Featuring AI Agent-Native OS, execution infrastructure is becoming more advanced. Model performance and infrastructure development are accelerating in parallel. The proposal to adjust the pace is a response to this acceleration. Differences in outlook within the industry may also surface going forward.

Rebuttal from Trump and Johnson

Trump and Johnson expressed negative views of the slowdown proposal. They argued that maintaining U.S. leadership should take priority. Underlying their stance is a sense of crisis that China could seize the lead. In an interview with the Financial Times, Trump said:

Look, we’re leading China in AI … and, frankly, I want to keep it that way, because whoever wins AI, wins.

He expressed the view that the United States is ahead of China in AI. He made clear his intention to maintain that lead. His understanding is that the outcome of AI will determine national victory or defeat.

Johnson answered questions from Jake Tapper on a CNN program. He said hasty regulation would pose a security threat. The original wording is as follows.

national security threat

He warned against rushed moves to strengthen regulation. He said that if Congress holds an emergency session to regulate, it will lose the race. The original statement is as follows.

If Congress just races in and does some sort of emergency session to try and regulate AI, we will lose the race to China.

He argued that fear-driven responses should be avoided. He did not deny the intent of regulation itself. His argument is that the wrong timing and approach would invite defeat.

The Clash Between Security and Regulation

This exchange is an intersection of technological and national security arguments. Amodei and others argued for the need to curb growing risks. Trump and others see the loss of advantage as the danger. The threats envisioned by each side differ.

U.S. AI policy has centered on export controls and investment restrictions. Restrictions on the supply of high-performance semiconductors are one example. The usual approach has been to maintain the pace of domestic development. The slowdown proposal challenges this premise. The administration’s backlash reflects policy continuity.

The role of Congress is also in focus. Johnson is cautious about emergency legislative action. His position is that regulation without sufficient deliberation would harm competitiveness. There is the reality that legal systems cannot keep up with the pace of technological change. Caution against hasty legislation remains strong.

Meanwhile, there are limits to voluntary slowdowns by companies. Because under a competitive structure, there is an incentive to break ranks. A slowdown without a common framework is difficult to sustain. Ensuring effectiveness requires a verifiable mechanism. The intersection of policy and voluntary measures is now in question.

The Contest Between Development Speed and

U.S.-China Competition

AI competition between the U.S. and China will not be decided by model performance alone. Computing resources, talent, and power infrastructure will determine the outcome. Speed of implementation in applied areas is also important. A pause in development would have multilayered effects. Simple comparisons of superiority are difficult.

The moves of Chinese companies are also being closely watched. Capability gains through distillation and domestic production of semiconductors have been reported. Moves to circumvent the U.S. regulatory net have also been observed. A change in pace will not immediately alter rankings. Even so, the administration views any delay itself as a loss.

The industry’s concern lies in the opacity of capabilities. High-performance models can exhibit unpredictable behavior. The development of evaluation methods has not kept pace with development speed. Ensuring safety requires time and resources. Balancing speed and verification is the challenge.

The future focus is on whether concrete measures will emerge. The details of Amodei’s proposal remain subject to debate. The administration has not presented alternative safety measures either. The confrontation remains at the level of principles. Attention is on a shift to practical discussions.

Editorial Opinion

In the short term, the clash between slowdown arguments and competition-first arguments could stall congressional deliberations. Submission of regulatory bills is likely to face postponement or revisions. Companies are expected to strengthen their statements on voluntary safety measures.

In the long term, whether a framework balancing speed and safety exists will determine competitiveness. If evaluation criteria and verification methods are developed, the reliability of development will increase. An international cooperation framework could also emerge.

As a question from the editors, whether competition without a slowdown can ensure safety remains in question. Whether criticism of overreaction is valid needs verification. What conditions would allow both maintaining speed and curbing risks is in question.

References

• “Trump and Mike Johnson think the AI industry is overreacting”, by Terrence O’Brien — The Verge, 2026-09-13T19:41:48.000Z (ARR)
• Source URL: https://www.theverge.com/ai-artificial-intelligence/994441/trump-mike-johnson-ai-industry-overreacting