Viral Wire

Congress Probes Airbnb's Use of Chinese AI Models Over Security Fears

Airbnb CEO defends Alibaba's Qwen model, but lawmakers flag data risks.

Deep Dive

US House committees on China and homeland security have launched a probe into Airbnb's use of Alibaba's Qwen large language model for its customer service chatbot. Airbnb CEO Brian Chesky defended the practice in a Bloomberg TV interview, stating the company uses open-source models (including US ones) and that open-source models do not have access to user data. He rejected concerns about Chinese firms accessing American data, calling the worries a misunderstanding. The committee chairs cited national security and data-security implications, arguing that Qwen is subject to China's ideological conditioning mandates, introducing structural vulnerabilities.

This probe reflects escalating US-China AI competition and a widening gap between Silicon Valley pragmatism and Washington's national security instincts. A new bipartisan bill, the U.S. Tech PATH Act, aims to counter Chinese low-cost tech dominance by streamlining procurement of American cyber technologies by allied governments, authorizing $500 million over five years. Meanwhile, data from OpenRouter shows Chinese open-source LLMs' global share surged from 1.2% in late 2024 to nearly 30% in just months of 2025, underscoring the economic pressures driving adoption of cheaper alternatives.

Key Points
  • Airbnb uses Alibaba's open-source Qwen LLM for its customer service chatbot; CEO Chesky claims no data is shared with Chinese companies.
  • House committee chairs cite national security concerns, arguing Qwen is trained under China's censorship regime and introduces hidden vulnerabilities.
  • A new bipartisan bill (U.S. Tech PATH Act) proposes $500M to help allied nations procure American tech instead of cheaper Chinese alternatives.

Why It Matters

Highlights the clash between cost-driven AI adoption and national security, affecting how US companies choose models.