A recent joint advisory issued by the National Security Agency (NSA), the FBI, and the Cybersecurity and Infrastructure Security Agency (CISA) has sparked concerns regarding the security of American artificial intelligence. The advisory alleges that Chinese AI developers are systematically extracting capabilities from US-based frontier models. Anthropic, a prominent US AI lab, has supported these claims, suggesting that Chinese firms may be secretly utilizing their Claude model to bridge the technological gap between the two nations.

The core of the allegation is that by leveraging existing US models, Chinese developers can bypass the significant time and financial investment required to train high-level AI from scratch. This practice, if confirmed, would allow these firms to reduce their own requirements for computing power and electricity. While the US government frames this as a national security issue involving the theft of intellectual property and strategic technological advantages, the allegations have created significant uncertainty within China’s tech sector. The situation highlights the ongoing geopolitical tension surrounding AI development, as Washington seeks to restrict the flow of advanced technology to China through export controls and security advisories. Currently, the specific methods used to access these models remain a subject of investigation, and the full extent of the alleged unauthorized usage has not been independently verified.