US Commerce Secretary Howard Lutnick Sounds Alarm On AI Threats, Says We 'Can't Let It Get Overrun By Chinese,' Calls For Industry-Led Security Audits Of Companies Like DeepSeek

In a stark warning on AI policy and national security, Commerce Secretary Howard Lutnick urged tighter controls on open-source AI models and called for American industry leaders to take the lead in securing the future of artificial intelligence.

What Happened: During a conversation on the All-In Podcast with Chamath Palihapitiya and David Friedberg that was posted earlier this week, Lutnick addressed the growing concern around foreign influence in AI, particularly from China — and stressed the need for a new framework to evaluate and secure AI systems developed outside the U.S.

Lutnick specifically cited open-source AI models such as DeepSeek expressing concern over their presence in the U.S. market.

"I don’t think we should be having apps in America and I don’t think we should have their website in America because they all go back home," Lutnick said, referring to companies based in China. "But it's open source, and I want our American companies, including college students, to be able to download it and build on it."

While Lutnick expressed support for innovation, he warned about potential national security risks embedded in AI systems. To address the issue, Lutnick proposed a security evaluation model driven by the private sector rather than traditional government regulation.

"We're going to have security evaluation and say if the security evaluation model says that this is a good model, then people can download it," he said. "But it's got to go through the industry, and I want it to feel and smell like what we're good at."

“So what I want to do is I’m going to embrace what you guys know. You guys used to do product evaluations. So let’s do security now," he stated.

Adding, “And say your industry, and you can’t let it get overrun by Chinese, because what happens is if there’s a policy, all of a sudden 100,000 people from China come in and they say they’re John Smith and Todd Peterson, but they’re not. And then you think the vote is this way, and it’s easily manipulated. So we have to be very careful.”

Subscribe to the Benzinga Tech Trends newsletter to get all the latest tech developments delivered to your inbox.

Why It's Important: Lutnick's remarks come amid intensifying concerns in Washington over China's role in the global AI arms race.

As the U.S. implements export controls on advanced Nvidia Corporation NVDA chips and places guardrails around where AI training can occur, officials are grappling with how to secure the open-source ecosystem.

Following the release of DeepSeek's R1 model in January 2025, investors offloaded Nvidia shares, causing the company’s market capitalization to decline by $600 billion. This led to a temporary loss of nearly 20% in Nvidia CEO Jensen Huang's net worth.

Earlier this month, Chinese start-up Qingcheng.AI unveiled a new AI framework that could reduce reliance on Nvidia's chips for AI model inference, marking the latest effort by Chinese tech companies to achieve technological self-sufficiency amid U.S. export restrictions.

Billionaire investor Ray Dalio has also highlighted China's dominance in the application and usage of AI technology while admitting that the U.S. was "uniquely competitive" in inventing, particularly the best chips in the world.

Amid the intensifying Chinese AI buzz, Cathie Wood-led Ark Invest increased its stake in Baidu Inc. BIDU.

Check out more of Benzinga’s Consumer Tech coverage by following this link.

Read Next: