Rep. Josh Gottheimer said Friday that he plans to introduce two bipartisan bills aimed at addressing national security and public safety risks posed by advanced artificial intelligence systems. Gottheimer, a Democrat from New Jersey and co-chair of the House Democratic Commission on Artificial Intelligence, announced the proposals at Fairleigh Dickinson University. He is working with Republican lawmakers on both measures as Congress weighs how to regulate increasingly powerful AI models. The American AI Security Act, co-sponsored by Rep. Mike Lawler, a New York Republican, would establish a mandatory federal review before advanced or “frontier” AI models could be released. The National Security Agency would lead the review, which would last 30 days and could be extended once for an additional 30 days. “Right now, there are no mandatory requirements for pre-deployment testing of frontier AI models,” Gottheimer said. “That doesn’t work in banking, it doesn’t work in food safety, it doesn’t work in medicine and it sure as hell shouldn’t work with the technology that experts say could help build a bioweapon or take down our power grid.” The proposal would put the federal government directly in the position of evaluating high-end AI systems before deployment. Gottheimer has argued that voluntary safeguards and industry-led assessments do not go far enough, including a provision in the bipartisan FRONTIER Act that would place independent safety assessors inside major AI companies. The New Jersey congressman’s plan generally follows a vetting process created by the Trump administration in response to cybersecurity concerns involving new models developed by companies including OpenAI and Anthropic. The administration has not publicly disclosed many details of how that system operates. Gottheimer also said he plans to introduce the China FIREWALL Act with Rep. Nick LaLota, a New York Republican. That legislation would prohibit Chinese-developed AI models with publicly available “open weights” from being used on government-issued devices. It would also bar federal agencies from purchasing software that depends on those models. The proposals come as Gottheimer takes a prominent role in shaping House Democrats’ approach to AI. House Minority Leader Hakeem Jeffries selected him and two other lawmakers to lead the party’s work on the technology. If Democrats win control of the House in the November midterm elections, that commission could help set the chamber’s legislative priorities on AI. Neither bill had been formally introduced as of Friday. Their prospects will depend on congressional action after the election, as well as negotiations over how to balance national security protections, innovation and the public’s ability to scrutinize powerful technology.