The United States and China are engaged in an AI competition driven by economic and strategic interests, but the increasing capability of AI raises concerns about human control. Experts suggest the current framework of an existential race is flawed and advocate for cooperation on safety standards.
The intense US-China competition in AI development, driven by national security and economic ambitions, could lead to unchecked advancements in autonomous systems. Experts warn that without cooperation on safety standards, the race may prioritize speed over human control, increasing risks from both state-sponsored and non-state actors.
The United States and China are engaged in a high-stakes competition over artificial intelligence, viewing it as a critical driver of economic, scientific, and strategic power. This dynamic creates a dilemma, as neither nation has a strong incentive to slow down AI development, despite growing concerns about the implications of increasingly capable and autonomous AI systems. The question of maintaining human control over AI has become paramount.
Recent discussions, including a summit between US President Donald Trump and Chinese President Xi Jinping, have brought this tension to the forefront. President Trump has indicated a preference for keeping AI development at its current pace without introducing new regulatory guardrails. This stance is echoed by comments from Treasury Secretary Scott Bessent, who emphasized the existential importance of the US not falling behind China in AI, stating that "If they were to pull ahead of us on AI, then nothing else matters."
However, an essay by Alvin W. Graylin, prepared for the Quincy Institute for Responsible Statecraft, challenges the prevailing narrative of an existential AI race. Graylin argues that the assumptions underpinning this race framework are flawed. He contends there is no single "finish line" for AI innovation, such as artificial general intelligence (AGI) or artificial superintelligence (ASI), that would grant a decisive and permanent strategic advantage to either nation. Capability gaps between the US and China in AI technology tend to be short-lived, typically lasting only a few months, and there is no "competitive moat" to secure a lasting lead for either country.
Furthermore, the essay highlights that the most significant AI security threats may not solely come from large, nation-state-sponsored models, but from smaller, specialized AI models that could be exploited by non-state rogue actors for malicious purposes, such as deploying biological, chemical, or cyber weapons. The current focus on US-China conflict, according to Graylin, hinders the necessary global cooperation to address these threats. The essay proposes that the summit between Trump and Xi presents an opportunity to shift from a model of existential competition to one of cooperation on shared harm and safety standards, advocating for the rejection of "national security exemptions" for AI labs and for measures to protect American workers from automation-induced displacement.
Pick the topics you care about. Get only what matters, on your cadence.