U.S. Lawmakers Propose Ban on Chinese AI in Federal Agencies

A group of U.S. lawmakers has introduced new legislation aimed at restricting the federal government’s use of artificial intelligence developed in countries considered geopolitical adversaries.

The bill, titled the “No Adversarial AI Act,” targets AI technologies originating from China, Russia, Iran, and North Korea.

The legislation was introduced in the House of Representatives by Republican John Moolenaar of Michigan and Democrat Raja Krishnamoorthi of Illinois.

Both lawmakers serve as leaders of the House Select Committee on the Chinese Communist Party. In the Senate, Senators Rick Scott of Florida and Gary Peters of Michigan are spearheading a companion version of the bill.

Key Restrictions and Oversight

If passed, the proposed law would:

• Ban federal agencies from acquiring or deploying AI models developed in the named countries.
• Require the Federal Acquisition Security Council to maintain and regularly update a list of such AI systems.
• Allow for exemptions only through approval by Congress or the Office of Management and Budget.
• Establish a removal process for AI models proven not to be under foreign adversarial control.

DeepSeek Under Scrutiny

The bill comes amid heightened scrutiny of foreign AI developers, particularly China-based DeepSeek. In January, the company announced that it had built an AI system with capabilities similar to ChatGPT, but at a lower cost.

However, U.S. intelligence assessments later linked DeepSeek to support for China’s military and intelligence services, as well as access to high-end Nvidia chips. Several federal agencies and U.S. companies have already restricted or banned the use of DeepSeek products.

The Trump administration has also reportedly considered formal action to limit the use of DeepSeek’s technology on government devices.

National Security at the Forefront

Representative Moolenaar emphasized the bill’s intent, stating that the U.S. must “draw a clear line” by preventing hostile AI systems from infiltrating government operations. The goal, he noted, is to safeguard federal networks from potential foreign surveillance or interference.

The bill also has support from lawmakers, including Democrat Ritchie Torres of New York and Republican Darin LaHood of Illinois. It aims to establish a long-term federal strategy for managing the risks of using foreign-developed AI technologies in sensitive government systems.