China's AI Models Rapidly Advance, Challenging US Dominance Amid Export Controls

Key Intelligence

• ■New, inexpensive Chinese AI models like GLM 5.2 are demonstrating performance comparable to leading Western models from Anthropic and OpenAI, indicating rapid advancement in China's AI capabilities.
• ■China is actively circumventing US export controls, with entities like Meituan training massive AI models using tens of thousands of domestic chips.
• ■The effectiveness and strategy of US AI model export controls are under scrutiny, with some experts suggesting that ad hoc restrictions could inadvertently accelerate China's indigenous AI development.
• ■Western developers like Anthropic are actively working to close loopholes that allow Chinese entities access to their frontier AI models.
• ■Beyond model performance, challenges in AI development include organizational integration, improving trustworthiness, and making AI's decision-making processes more transparent.