A bipartisan coalition of lawmakers wants to stop U.S. agencies from acquiring or using artificial intelligence made by the Chinese Communist Party or other foreign adversaries.
The House Select Committee on the CCP created the No Adversarial AI Act and said Wednesday it has the support of Sens. Rick Scott, Florida Republican, and Gary Peters, Michigan Democrat.
Rep. John Moolenaar, the committee chairman, said AI is at the center of a new Cold War with China and added that the U.S. must stop CCP theft and subversion.
“From IP theft and chip smuggling to embedding AI in surveillance and military platforms, the Chinese Communist Party is racing to weaponize this technology,” the Michigan Republican said in a statement. “We must draw a clear line: U.S. government systems cannot be powered by tools built to serve authoritarian interests.”
The lawmakers’ proposal intends to create a public list of AI systems shaped by foreign adversaries. Federal agencies would be prohibited from adopting AI systems on the list except for certain cases such as research and counterterrorism, according to the committee.
The legislation also proposes to create a delisting process for companies that demonstrate they removed themselves from the influence of America’s adversaries.
Mr. Scott said lawmakers have “clear evidence” that China can have access to Americans’ data via AI systems and it is “absolutely insane” to permit federal agencies to use such systems.
Rep. Raja Krishnamoorthi, the committee’s ranking Democrat, said he co-sponsored the legislation to help establish a firewall between foreign adversaries’ AI and American-developed tools.
Alongside the Illinois Democrat and Mr. Moolenaar, the House legislation also has sponsorship from Reps. Darin LaHood, Illinois Republican, and Ritchie Torres, New York Democrat.
During a committee hearing on Wednesday, Mr. Krishnamoorthi said he’s also concerned about the emergence of artificial general intelligence, which performs at least as well as humans across all cognitive domains.
He said he is working on additional legislation to govern such powerful AI that has yet to emerge.
“Whether it’s American AI or Chinese AI, it should not be released until we know it’s safe — that’s why I’m working on a new bill, the AGI Safety Act, that will require AGI to be aligned with human values and require it to comply with laws that apply to humans,” Mr. Krishnamoorthi said during the hearing.
AI is a major emphasis for the select committee, which is reviewing export controls, ways to supply chain security for AI systems, and efforts to ensure U.S. innovation doesn’t power foreign surveillance or military systems.