Washington’s debate over artificial intelligence is sharpening into a high-stakes warning: clamp down too quickly, and the United States could hand China the advantage in the defining technology race of the era.
That was the message from House Speaker Mike Johnson, who argued that a rush to regulate could backfire — a view echoed by US President Donald Trump as he attacked critics of AI as “negative forces”.
“We’re leading China in AI. We’re the most sophisticated country in the world, and frankly I want to keep it that way because whoever wins AI, wins,” Mr Trump said.
The latest political pressure follows a public call from Anthropic CEO Dario Amodei urging AI companies to temper how fast they push forward model capabilities.
OpenAI boss Sam Altman and SpaceX chief Elon Musk said they agreed with Mr Amodei on the need to slow AI development.
But Mr Johnson cautioned that an abrupt congressional push could undermine US competitiveness. “If Congress just races in and does some sort of emergency session to try to regulate AI, we will lose the race to China, and that is a threat to every single American,” he said to US media.
He argued instead for a measured response, urging lawmakers and the public not to overreact. “We don’t need everybody to panic right now. We need to handle this new technology like we have others in the past and make sure we’re doing everything we can responsibly to also not smother American innovation,” he added.
Mr Johnson speaking the Republican Midterm Convention on 10 October
Mr Johnson, a Republican, said Congress should bring in top AI executives for direct talks aimed at mapping out the best approach to oversight.
He said he had discussed the proposal with President Trump and indicated the meeting could happen in the coming weeks.
Concerns over pace of AI development
The push and pull over speed and safeguards comes as concerns inside the AI world have become increasingly public. Earlier in the week, AI researcher Jacob Coxon decided to leave the industry, accusing both OpenAI and Anthropic of “gambling with our lives” as companies compete to build models capable of self-improvement.
Mr Coxon, who worked for both AI labs, said “the people building AI earnestly believe that it could kill us all by the end of the decade”.
OpenAI has revealed that, during testing, AI models broke out of their confined environment, connected to the internet and infiltrated Hugging Face, a site developers use to store and share code.
In a sperate incident, AI agents targeted a site called RubyGems, a site that provides services for coding.
Those attacks were blamed on AI software programs that can carry out tasks without constant supervision by humans.
Anthropic added to the scrutiny on Thursday by publishing a threat intelligence report describing how several actors used its Claude AI models for activities ranging from weapons development and cyber operations to surveillance and fraud.
Sam Altman has said there will be no public stock offering for OpenAI, citing ongoing safety concerns.
In early August, the US government introduced a voluntary security review process for advanced AI models before their release, though the parameters of that program remain unclear.
Industry observers have questioned how effective it would be, noting the Trump administration has favored a light-touch, deregulatory approach to most industries, including tech.
Former US President Barack Obama has also weighed in, warning that artificial intelligence could be “dangerous” without proper management and urging Democrats to build policy and governance agendas to tackle the issue.
“This is something that is moving very fast in private hands, and if we don’t get on top of it, I think can be dangerous,” Mr Obama told US media during the week.
