Voices including major AI developers such as Anthropic and OpenAI are calling for slowing the development of advanced AI models to put safety and control mechanisms in place, but governments are instead showing a negative stance toward the idea of throttling back speed.

Governments typically seek to regulate and companies are used to calling for deregulation, but AI development is producing the opposite picture.

U.S. President Donald Trump has brushed aside the need to strengthen federal regulation of AI and branded safety concerns a scam. He also warned AI regulation would bring down AI companies and hand China a competitive advantage.

In a post on social media platform Truth Social on Sept. 14 local time, Trump said, "The only safeguard AI needs is a strong and smart president, and America has exactly that." He added, "There is an evil plot targeting AI and data centers, and the only country that welcomes it is China."

In a separate post, he pushed back by asking why industry leaders were calling for regulation and said such demands could drive the industry into bankruptcy.

With midterm elections in November approaching, Trump's stance diverges from broader public opinion. A Gallup poll in March showed nearly half of respondents strongly opposed construction of local AI data centers, and nearly 40 percent of Republicans also opposed it. As a result, AI data centers are becoming a political burden for both parties ahead of the election, The Wall Street Journal reported.

Recently, executives at major U.S. AI companies, including Anthropic CEO Dario Amodei (다리오 아모데이), have called for slowing AI development out of concern about risks. Amodei urged moderating the pace of AI development in a blog post and stressed international cooperation on the issue. "Even buying 1 to 2 years before model capabilities reach dangerous levels can greatly reduce the risk of a serious accident," he said.

OpenAI CEO Sam Altman (샘 알트먼) and Elon Musk have also moved in step with Amodei. Altman warned on social media X (Twitter) that "humanity could lose control of the future to AI" and said he "welcomes the introduction of consistent safety standards at the federal level."

Jacob Coxon (제이콥 콕슨), a former Anthropic researcher, submitted his resignation and conveyed concerns that Anthropic and its rivals are competing to develop uncontrollable systems.

Even so, the U.S. government is skeptical about regulating AI the way the industry wants. Vice President Vance said, "The government must regulate intelligently," but he is skeptical about introducing additional safety mechanisms as companies are demanding. "The way frontier AI companies are begging the government to regulate them looks like a Trojan horse," he said. House Speaker Mike Johnson said Trump is expected to meet AI companies at the White House this week to discuss solutions.

Cooperation among specific countries is key to moderating the pace of AI development, but that also does not look easy for now. In particular, there is a view that it is realistically difficult for the United States and China, which are competing in the global AI development race, to become "one team" on regulation. Amodei cited as the biggest dilemma of his proposal to slow advanced AI development the strong likelihood that China will not join.

Amodei, Anthropic CEO: "The biggest dilemma of AI slowdown is China."

The Chinese government has already pushed back against calls from the U.S. AI industry to slow development. CNBC reported that China's foreign ministry said calls to slow AI development could instead hinder the global AI governance process.

China dismisses U.S. AI industry's slowdown calls: "Fearmongering disrupts AI governance."

In the venture capital industry, which is tied to AI developers through shared interests, many see the slowdown push as an overreaction. David Sacks (데이비드 색스), a venture investor and co-chair of the U.S. President's Council of Advisors on Science and Technology, also made clear he opposes the slowdown calls raised by Anthropic and OpenAI.

David Sacks criticises Anthropic-OpenAI slowdown calls: "They can just not build it."

Michael Burry (마이클 버리), known as the real-life figure behind "The Big Short" and for predicting the 2008 subprime mortgage collapse, dismissed the AI safety warnings raised by OpenAI and Anthropic as selfish IPO hype. "The demand to slow frontier AI development is really intended to protect incumbent companies as competitors catch up," he said, and claimed they are inflating the idea that their technology could become dangerous because it is too powerful.