House bipartisan bill directs NSA to create 'AI security playbook' amid Chinese tech race

FIRST ON FOX – Rep. Darin LaHood, R-Ind., is introducing a caller measure Thursday imploring the National Security Administration (NSA) to make an "AI information playbook" to enactment up of threats from China and different overseas adversaries. 

The bill, dubbed the "Advanced AI Security Readiness Act," directs the NSA’s Artificial Intelligence Security Center to make an "AI Security Playbook to code vulnerabilities, menace detection, cyber and carnal information strategies, and contingency plans for highly delicate AI systems." 

It is co-sponsored by House Select Committee connected China Chairman Rep. John Moolenaar, R-Mich., Ranking Member Rep. Raja Krishnamoorthi, D-Ill., and Rep. Josh Gottheimer, D-N.J. 

LaHood, who sits connected the House Intelligence Committee and the House Select Committee connected China, told Fox News Digital that the legislative proposal, if passed, would beryllium the archetypal clip Congress codifies a "multi-prong attack to guarantee that the U.S. remains up successful the precocious exertion contention against the CCP." He said the measure volition amended export power mechanisms – including for chips and precocious capableness spot manufacturing – support covered AI technologies with a absorption connected cybersecurity, and bounds outbound concern to firms straight tied to the Chinese Community Party oregon China's People's Liberation Army. 

CHINESE BIOWEAPON SMUGGLING CASE SHOWS US 'TRAINS OUR ENEMIES,' 'LEARNED NOTHING' FROM COVID: SECURITY EXPERT

Rep. Darin LaHood, R-Ill., leaves the House Republican Conference gathering astatine the Capitol Hill Club successful Washington, D.C., connected Tuesday, May 17, 2022. (Bill Clark/CQ-Roll Call, Inc via Getty Images)

"We commencement with the premise that China has a program to regenerate the United States. And I don't accidental that to scare radical oregon my constituents, but they person a program to regenerate the United States, and they're moving connected it each azygous day. And that entails stealing information and infiltrating our systems," LaHood told Fox News Digital. "AI is the adjacent frontier connected that. We pb the satellite successful technology. We pb the satellite erstwhile it comes to AI. But what this measure volition bash volition again marque definite that things are done the close mode and the close way, and that we're protecting our assets and promoting the existent exertion that we person successful our country." 

LaHood pointed to grounds uncovered by the committee that helium said shows the CCP's DeepSeek utilized amerciable distillation techniques to bargain insights from U.S. AI models to accelerate their ain exertion development. He besides pointed to however China allegedly smuggled AI chips done Singapore intermediaries to circumvent U.S. export controls connected the technology. 

"As we look at, ‘How bash we triumph the strategical competition?’ I deliberation astir experts would accidental we're up successful AI close present against China, but not by much. It is simply a abbreviated lead," LaHood told Fox News Digital.

He said helium is assured the measure volition enactment the U.S. "in the champion presumption to support our assets present and marque definite that we're not shipping things that shouldn't spell to AI that let them to triumph the AI contention successful China."

"Whoever wins this contention successful the future, it's going to beryllium captious to aboriginal warfare capabilities, to, obviously, cybersecurity," LaHood continued. "And then, whoever wins the AI contention is going to output truly unwavering economical power successful the future. And truthful we're assertive successful this measure successful presumption of targeting those areas wherever we request to support our AI and our companies present successful the United States, some connected the commercialized broadside and connected the authorities side, to enactment america successful the champion presumption possible." 

National Security Agency Director Gen. Timothy Haugh speaks during a Senate Committee connected Intelligence Hearing connected March 25, 2025 successful Washington, D.C. (Andrew Harnik/Getty Images)

The legislative connection calls connected the NSA to make a playbook that identifies vulnerabilities successful AI information centers and developers producing delicate AI technologies with an accent connected unsocial "threat vectors" that bash not typically arise, oregon are little severe, successful the discourse of accepted accusation exertion systems." The measure says the NSA indispensable make "core insights" successful however precocious AI systems are being trained to place imaginable interferences and indispensable make strategies to "detect, forestall and respond to cyber threats by menace actors targeting covered AI technologies." 

AMAZON ANNOUNCES $20B INVESTMENT IN RURAL PENNSYLVANIA FOR AI DATA CENTERS

The measure calls connected the NSA to "identify levels of security, if any, that would necessitate important involvement" by the U.S. authorities "in the improvement oregon oversight of highly precocious AI systems." It cites a "hypothetical inaugural to physique covered AI exertion systems successful a highly unafraid authorities environment" with definite protocols successful place, specified arsenic unit vetting and information clearance processes, to mitigate "insider threats." 

Though not straight related, the authorities is being introduced a week aft FBI Director Kash Patel sounded the alarm connected however the CCP continues to deploy operatives and researchers to "infiltrate" U.S. institutions. Patel laid retired the hazard successful announcing that 2 Chinese nationals were charged with smuggling a imaginable bioweapon into the U.S. 

LaHood said that lawsuit further highlights "the level of penetration and sophistication that the CCP volition prosecute in," but helium added that his measure focuses connected putting a "protective layer" connected U.S. AI tech and "restricting outbound concern to China." He pointed to however the CCP besides has bought up farmland astir strategical U.S. nationalist information locations, peculiarly successful Montana, North Dakota and South Dakota. 

Rep. Darin LaHood, R-Ind., said China's extremity is to regenerate the United States. (ADEK BERRY/AFP via Getty Images)

"If everything was an adjacent playing field, and we were each abiding by the aforesaid rules and standards and ethical guidelines, I person nary uncertainty the U.S. would triumph [the AI race], but China has a inclination and a past of playing by a antithetic acceptable of rules and standards," LaHood said. "They cheat, they steal, they instrumentality our intelligence property. Not conscionable my opinion, that's been factually laid out, you know, successful galore antithetic instances. And that's the crushed wherefore we request to person a measure similar this." 

The measure comes arsenic the Trump medication has been pushing to bolster artificial quality infrastructure successful the United States, and large tech companies, including Amazon, Nvidia, Meta, OpenAI, Oracle and others, person made large investments successful constructing AI-focused information centers and enhancing U.S. unreality computing. Last week, Amazon announced a $20 cardinal concern successful constructing AI information centers successful agrarian Pennsylvania. It followed a akin $10 cardinal investment successful North Carolina. 

CLICK HERE TO GET THE FOX NEWS APP 

In precocious May, the NSA's Artificial Intelligence Security Center released "joint guidance" connected the "risks and champion practices successful AI information security." The recommendations see implementing methods to unafraid the information utilized successful AI-based systems, "such arsenic employing integer signatures to authenticate trusted revisions, tracking information provenance, and leveraging trusted infrastructure." The halfway said its guidance is "critically applicable for organizations – particularly strategy owners and administrators wrong the Department of Defense, National Security Systems, and the Defense Industrial Base – that already usage AI systems successful their day-to-day operations and those that are seeking to integrate AI into their infrastructure."