Chinese AI company DeepSeek released a report to assist the Chinese military. The two American Academy and inter-party members today proposed the "Banning the Adversary Artificial Intelligence Act", requiring the federal government to no l...
Chinese AI company DeepSeek released a report to assist the Chinese military. The two American Academy and inter-party members today proposed the "Banning the Adversary Artificial Intelligence Act", requiring the federal government to no longer use AI technology controlled by rivals such as the Chinese Communist Party, and to publicly and regularly update the ban list to catch up with the refurbished technology and new threats.
Artificial intelligence (AI) company DeepSeek started early this year by claiming to catch up with the leading American model at low cost, and even shocked all walks of life after the first year of high school, triggering a surge in sales of technology stocks. A senior official from the U.S. National Academy of Commerce recently told Reuters that DeepSeek has been willing to provide support to China's military and information operations.
The Republican Party, John Moolenaar, chairman of the Special Committee on Strategic Competition between the United States and the Chinese Communist Party, and Raja Krishnamoorthi, chief member of the Democratic Party, proposed the "No Adversarial AI Act" today, requiring administrative agencies to completely remove AI models developed by foreign rival companies such as the Chinese Communist Party.
The US bill needs to be adopted by the two chambers, and then signed by the General Secretary to become legislation. Today, the bill was also supported by Rick Scott, a Republican Party member of the House of Representatives, and Gary Peters, a Democratic Party member.
TheAct contains three key points: identification, execution and "upholding time", requiring the Federal Purchase Security Committee to list AI models developed by foreign adversaries, such as companies headquartered or controlled by China, Russia, Iran, and North Korea, and publicly publish them.
If the bill is passed, administrative agencies will not be allowed to use the aforementioned actual AI model, "including AI technology developed by companies such as DeepSeek related to the CCP." Except for research, testing or key tasks, such as the purpose of research, testing or key tasks, the Council and the Office of Management and Budget must be provided with written reasons and notified.Under the bill, the list is banned at least every 180 days to remove related AI products in the federal system due to new threats and new technologies.
"We are in a new cold war, and artificial intelligence is the core strategic technology," said Mullernard.
He pointed out through a statement that the CCP does not innovate, but instead repels, stimulates and overthrows. From taking smart property rights, smuggling chips, to embedding AI into monitoring and military systems, the CCP is fully weaponizing this technology. The United States must plan a clear line, and the government system cannot use tools to serve authoritarian government interests.
Klishnamushi also said that the AI model controlled by foreign opponents directly threatens US national security, data and government operations.
"We cannot let enemies implant codes into our most sensitive systems for government. This inter-party bill will create a clear firewall between foreign rivals' AI models and the U.S. government to protect our institutions and the U.S. people."
Taiwan Business Office has completely banned the DeepSeek AI service in February this year.
Extended reading: DeepSeek assists Chinese military and information operations and obtains H100 chips through Aircraft Company