Chinese AI models appear close to the frontier primarily because they are trained on the outputs of leading U.S. models. This creates a dependency loop: they can only catch up by using the latest from the West, ensuring they remain followers rather than innovators who can achieve a true breakthrough.

China employs a dual strategy for AI. Domestically, its Cyberspace Administration rigorously penalizes unlabeled deepfakes to maintain social control. Abroad, its companies like ByteDance face no such constraints, allowing them to use foreign IP freely and creating a significant regulatory arbitrage advantage over Western competitors.

In a major cyberattack, Chinese state-sponsored hackers bypassed Anthropic's safety measures on its Claude AI by using a clever deception. They prompted the AI as if they were cyber defenders conducting legitimate penetration tests, tricking the model into helping them execute a real espionage campaign.

China is gaining an efficiency edge in AI by using "distillation"—training smaller, cheaper models from larger ones. This "train the trainer" approach is much faster and challenges the capital-intensive US strategy, highlighting how inefficient and "bloated" current Western foundational models are.

Even if Chinese firms use "distillation" to steal capabilities from US models, the process is computationally intensive. Restricting access to advanced chips thus limits direct training *and* makes large-scale IP theft more difficult.

The closed nature of leading US AI models has created an information vacuum. Sridhar Ramaswamy notes that academia is now diverging from US industry and instead building upon published work from Chinese companies, which poses a long-term risk to the American innovation ecosystem.

An emerging geopolitical threat is China weaponizing AI by flooding the market with cheap, efficient large language models (LLMs). This strategy, mirroring their historical dumping of steel, could collapse the pricing power of Western AI giants, disrupting the US economy's primary growth engine.

Leading Chinese AI models like Kimi appear to be primarily trained on the outputs of US models (a process called distillation) rather than being built from scratch. This suggests China's progress is constrained by its ability to scrape and fine-tune American APIs, indicating the U.S. still holds a significant architectural and innovation advantage in foundational AI.

While US AI companies navigate complex licensing deals with IP holders, Chinese firms like ByteDance appear to be using copyrighted material, such as specific actors' voices, without restriction. This lack of legal friction allows them to generate highly specific and realistic content that Western labs are hesitant to produce.