China Reportedly Refuses Call to Limit AI Use in Nuclear Arms

A policy adopted by the Biden administration that limits the use of AI for nuclear weapons has not been embraced by the Chinese government. 

A second White House official and White House National Security Council tech director, Tarun Chhabra, has announced that the White House will soon issue a national security memorandum detailing the use of artificial intelligence (AI) by different US government agencies, including the Pentagon’s development of new nuclear weapons and hypersonic missiles. 

Due to AI’s growing influence as a weapon in national security strategies, the United States maintains that other nations should adopt this approach.

Keeping the lines of communication open in the U.S.-China AI negotiations is essential to the White House, as it will enable them to address the risks and find ways to make the technology safe. In their AI discussions, the US and China hope to establish a “risk hierarchy” for military AI and share details about testing, evaluating, validating, and verifying systems. The United States also wants China to promise that human beings should be kept informed about any developments concerning the command and control of nuclear weapons.

The specific application of AI in nuclear command and control by China remains a mystery. Reportedly still in operation today is the autonomous nuclear launch system known as “Dead Hand,” which the Soviet Union built during the Cold War. If the country’s leaders are murdered or crippled, the system, which is also called “Perimeter,” may autonomously fire long-range nuclear missiles.

The following memorandum on artificial intelligence will aim to maintain the United States’ technological leadership and promote responsible AI use, according to Maher Bitar, who is the National Security Council’s coordinator for intelligence and military policy and a deputy assistant to the president. The Department of Energy and national labs, the Department of Homeland Security, the government Bureau of Investigations (FBI), the Department of Justice (DOJ), and other government agencies will all be included in the memorandum.

Another area of worry is the potential for AI to be utilized by authoritarian regimes like China’s Communist Party to conduct internal surveillance and control systems in violation of human rights. AI might also track U.S. intelligence officers, which is a double-edged sword that could compromise both human rights and counterintelligence efforts to safeguard American citizens, assets, and government operations.