Johnson Rejects Broad AI Moratorium, Warns China Could Gain Advantage

House Speaker Mike Johnson opposed an AI development moratorium, arguing excessive regulation could surrender America's technological advantage to China.

U.S. House Speaker Mike Johnson has rejected calls for Congress to impose a broad moratorium on artificial intelligence development, arguing that slowing American AI progress could hand a strategic advantage to China. His position comes as Washington faces an increasingly difficult debate over how to regulate rapidly advancing AI systems without weakening the United States' technological position. Johnson has argued that AI is becoming a major component of national security and economic competitiveness and that allowing development to slow could create vulnerabilities if Chinese companies and institutions continue advancing at a faster pace. His comments came amid growing pressure from researchers and AI executives for stronger safeguards, particularly around frontier systems that could become capable of autonomous reasoning, cyber operations or other activities that create risks difficult for existing regulatory structures to manage. Johnson has indicated that he does not believe Congress should lead the response by imposing a sweeping development pause, instead favouring industry responsibility combined with targeted safeguards.

The debate has intensified because the United States is attempting to maintain leadership across several layers of the AI ecosystem simultaneously. American companies currently dominate much of the frontier-model industry, while U.S. firms also have major positions in cloud computing, advanced chips and AI infrastructure. Policymakers therefore increasingly view AI not simply as a commercial technology but as a strategic asset comparable in importance to previous transformational technologies. A development moratorium could reduce some immediate safety risks, but supporters of continued rapid development argue that stopping American companies would not necessarily stop competitors elsewhere. China, in particular, is investing heavily in domestic AI models, computing infrastructure and applications across government and industry. Johnson's argument reflects this geopolitical concern: if U.S. companies are restricted while foreign competitors continue operating, the United States could lose technological leadership and potentially become dependent on foreign AI systems. At the same time, critics argue that treating AI competition as an unrestricted race could encourage companies to prioritise speed over safety, creating risks that individual firms may be unable or unwilling to manage on their own.

Johnson has nevertheless acknowledged the need for some form of guardrails and has supported discussions involving major AI companies, independent auditors and greater transparency. That approach reflects an emerging middle ground between two competing positions in Washington. One side favours significantly stronger government oversight, independent testing and enforceable requirements for frontier systems. The other argues that excessive regulation would slow innovation and make American companies less competitive internationally. The challenge is designing safeguards that target genuinely dangerous capabilities without creating a regulatory structure so broad that it prevents useful innovation. The debate is likely to remain central to U.S. policy as increasingly capable models move from experimental systems into commercial and government applications. Johnson's position makes clear that national-security competition with China will be a major consideration in that debate. Rather than treating AI safety and technological leadership as separate questions, Washington is increasingly being forced to decide how both objectives can be pursued simultaneously. The outcome could shape the regulatory environment for the world's most powerful AI companies and determine whether the United States pursues a model based primarily on voluntary safeguards, targeted legislation or a substantially more interventionist system.