As President Donald Trump prepared to meet with Chinese President Xi Jinping, a top Pentagon official on Sept. 23 rejected calls for a slowdown in artificial intelligence (AI) development, saying “unliteral disarmament” in the AI race with Beijing would be a grave mistake.
“Artificial intelligence is accelerating quite rapidly, and maintaining U.S. dominance there is something that we need to focus on, and we need to continue to push for,” said Cameron Stanley, chief digital intelligence officer for the Department of Defense (DOD), rebranded as the Department of War by the Trump administration. “The U.S. tech sector is ahead. It’s going to stay ahead.”
“While there are risks with any technology that you’re integrating … unilateral disarmament is the wrong approach,” added Stanley, who said the Pentagon is moving ahead with AI development and that up to 2 million people a week are now using its GenAI.mil frontier AI platform.
Stanley’s comments at a Punchbowl News event came hours before Xi was scheduled to arrive in the United States and amid a continuing debate over AI safety and whether the U.S. should slow – and potentially regulate – the development of advanced AI systems.
The issue was supercharged by a Sept. 12 essay by Anthropic CEO Dario Amodei, in which he wrote: “We must slow the pace at which we improve the capabilities of AI models.”
Elon Musk, the CEO of xAI, and OpenAI CEO Sam Altman also backed calls to slow the development of advanced AI systems while researchers, regulators, and industry develop stronger safeguards.
In Congress, Sen. Bernie Sanders, I-Vt. on Sept. 23 introduced legislation he had previously highlighted that would pause advanced AI development until safeguards are in place. Sanders has called on Trump and Xi to negotiate an international agreement for a pause and to prohibit superintelligence.
But Trump has rejected the concerns about AI, describing opposition to the technology and data center development as a “SICK conspiracy” and arguing that China would benefit from U.S. restrictions on AI. Trump and Xi are considered likely to discuss AI safety in their talks.
Stanley voiced agreement with Trump in his remarks, saying there is “a big difference between the Chinese approach to AI and the U.S. approach to AI, and we have a great relationship with our frontier labs … that commercial dominance [is] something that we want to harness for our U.S. Department of War objectives and outcomes, and we’re integrating them into a lot of different workflows.”
Those workflows, he said, include an AI pilot program to try to accelerate defense contracting and other AI-fueled efforts to update intelligence, logistics, and overall warfighting.
As for GenAI.mil, Stanley called the platform “a huge, a huge win for the department. Before it was rolled out, we had a total of maybe 80,000 folks that were using AI across the entire department.”
“It’s an awakening across the entire department that data-driven decision-making is the right approach,” Stanley said. “Leveraging AI for the benefit of the American warfighter is the right thing to do … this could not be more important for the defense of our country.”