Policy

US Lawmakers Push to Ban Chinese AI Models from Federal Systems

Congressional report identifies national security risks from widespread adoption of Chinese-developed AI across government and critical infrastructure.

Omega Editorial · November 21, 2026 · 3 min read

Key takeaways

  • House Republicans released a report calling for bans on Chinese AI models in federal systems and critical infrastructure, expanding beyond existing DeepSeek restrictions
  • A seven-month investigation found US companies including Airbnb, Anysphere, and DoorDash have adopted Chinese AI models due to competitive pricing and capabilities
  • The report alleges Chinese developers used proxy networks and fraudulent accounts to extract capabilities from American AI systems through model distillation
  • Lawmakers propose creating an "AI bill of materials" to track model origins and establishing standards to identify which AI components are integrated into US systems
  • The recommendations have not been enacted into law and build on earlier warnings from NSA, CISA, and FBI about Chinese AI capability extraction efforts

Two Republican lawmakers have called for comprehensive restrictions on Chinese-developed artificial intelligence models in US government systems and critical infrastructure, warning that their growing adoption creates national security vulnerabilities.

House Homeland Security Committee Chairman Andrew Garbarino and House Select Committee on China Chairman John Moolenaar released an investigative report on November 21 containing 12 policy recommendations aimed at reducing American dependence on Chinese AI technology. The report, based on a seven-month investigation launched in April, examined how US companies are increasingly adopting Chinese AI models due to their competitive capabilities and significantly lower costs.

Proposed restrictions target government and infrastructure

The lawmakers urged the federal government to prohibit Chinese-developed AI models in government systems and federal contract work, remove them from sensitive functions within critical infrastructure, and prevent federal cybersecurity grants from purchasing such technology. Current restrictions focus primarily on DeepSeek, a Chinese AI company already excluded from federal civilian agency systems under a Department of Homeland Security order. The new recommendations would expand these prohibitions more broadly.

The investigation examined several American technology companies, including Airbnb, Anysphere, and DoorDash. According to the report, Airbnb has incorporated AI models from Chinese technology company Alibaba into customer support operations. Anysphere built its Cursor AI coding platform using a base model from China's Moonshot AI. DoorDash reported that several families of Chinese-developed AI models were authorized and deployed within its operations. The report did not establish that any cyberattacks or unauthorized disclosures resulted from these companies' use of Chinese models.

Why it matters

The report highlights a strategic vulnerability in US technology infrastructure: American companies are adopting Chinese open-weight AI models because domestic alternatives cannot match their price-performance combination. This creates potential dependencies that could be exploited for intelligence gathering or system manipulation, particularly as AI becomes embedded in critical infrastructure. The findings also reveal alleged intellectual property concerns, with federal investigations identifying instances where Chinese developers used proxy networks and fraudulent accounts to extract capabilities from American AI systems through model distillation techniques.

Recommendations beyond restrictions

Beyond banning Chinese models, the lawmakers proposed establishing standards to identify AI model origins and creating an "AI bill of materials" that would allow organizations to determine which models and components are integrated into their systems. Other recommendations included strengthening cybersecurity information-sharing, creating a dedicated intelligence channel for AI-related threats, and empowering the Cybersecurity and Infrastructure Security Agency to coordinate reporting of AI-related cyber incidents.

The report also called for a national initiative to accelerate development of American open-weight AI systems and make trusted US technology more accessible internationally. The lawmakers emphasized avoiding excessive regulation that could hinder American AI innovation while developing common standards with international allies.

The proposed restrictions have not been enacted into law. The findings build on September warnings from the National Security Agency, CISA, and FBI about large-scale efforts by China-based AI companies to extract capabilities from leading American models. Details were first reported by Anadolu Agency.

  • #chinese ai
  • #federal cybersecurity
  • #deepseek
  • #ai regulation
  • #critical infrastructure
  • #open-weight models

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy