Letting China pull ahead of the U.S. in the race for dominance over artificial intelligence could dramatically threaten the U.S. economy and national security.
That’s the warning that Horacio Rozanski, CEO of the global technology company Booz Allen Hamilton, offered in a wide-ranging interview on The Washington Times’ “Threat Status” podcast.
Mr. Rozanski, who joined the podcast from the 2025 Reagan National Defense Forum in California, outlined threats posed by advanced AI, especially when wielded by bad actors in the cybersecurity realm.
He also spoke of U.S.-China competition for leadership over the futuristic technology, asserting that the winner will hold greater influence around the world.
“This is a race,” Mr. Rozanski said. “And one country is going to win.”
He added three points are at play: “One, who has the best technology stack that will build the best models and everything else. Two, who’s driving adoption, because if you don’t adopt the technology, it’s not useful. And, three, who’s applying it most intelligently to national security.
“All three of those dimensions are really important, and as Americans we need to be working together and pushing for the U.S. to win across all three fronts,” said Mr. Rozanksi, who added that Booz Allen Hamilton is the largest provider of AI and cybersecurity to the U.S. federal government.
His comments coincide with Trump administration efforts to reduce barriers for U.S.-based AI companies for a technology that could help or hurt the economy, American culture and society.
President Trump most notably introduced an AI action plan in July titled “Winning the Race.”
Mr. Rozanski said the “administration is focused on this, and if you look at the action plan, I was very pleased to see some aspects of security in the adoption … because at the end of the day, if we don’t trust these [AI] models because they’re being corrupted, because they’re doing the wrong thing and so forth, we’re not going to use [them, and] if we don’t use them, we’re going to fall behind.”
On a separate front, the CEO emphasized risks associated with “AI being used by bad actors on the cyber front to corrupt networks, to get into companies and steal their data and their information, to create ransomware.
“That is right here, right now,” Mr Rozanski said, adding that “the level of power and the level of risk, we have not seen before. The world is not ready” for the cyber risks associated with AI.
Experts have increasingly sounded the alarm over AI’s capacity to serve as a force multiplier for malicious cyber actors. In the field of ransomware specifically, AI tools have let bad actors create convincing deepfakes of company executives and automatically generate new variations of malware to skirt proprietary cybersecurity software.
Some cybercriminals are based in China, with ties to the Chinese Communist Party and Chinese intelligence agencies.
The U.S.-based AI developer Anthropic recently revealed that it detected a disruption operation using the company’s Claude AI tools. The operation targeted over 30 global entities, including tech companies, financial institutions, chemical manufacturing companies and government agencies. Anthropic reported with “high confidence” that the operation was orchestrated by a Chinese state-sponsored group.
Mr. Rozanski said similar operations and their fallout will likely worsen, especially as China looks to move its AI operations into space.
He warned that AI-powered attacks could soon occur in space, potentially threatening Western GPS infrastructure and destabilizing the U.S. economy.
“Every aspect of the economy is tied to space and certainly national security,” Mr. Rozanski said. “Every ATM in the country leverages a GPS signal to figure out what time it is, which is tied to a satellite. And if the ATM doesn’t have confidence in what time it is, it cannot give you money because it doesn’t know when your last deposit was and where we are in time. So if you disrupted that constellation, all of a sudden the banking system gets disrupted.”
With regard to who’s winning the AI race overall, Mr. Rozanski said: “Now, the U.S. is ahead. But one of the things that China is working on is a national data center in space.
“They’re building an AI constellation, where they’re going to be to compute in space, which has some advantages around energy and heat dissipation.”
Other experts say China’s effort, known as the “Three-Body Computing Constellation,” is still in development and may be operational by the end of 2026. The network of 2,800 satellites could have a massive processing output of 1 quintillion operations per second, according to some estimates.
Beijing has rapidly expanded its space capabilities the past few years, with over 1,300 satellites in orbit and more on the way. Security experts have warned that if the U.S. falls behind China in launch capabilities and fielded technology, it faces serious risks to national security.
Pentagon leaders voiced such concerns this week at the Space Force Association’s Spacepower 2025 conference in Orlando, Florida. Air Force Secretary Troy Meink said Beijing has made major advances in launch capability and is poised to surpass the U.S. in space.
“It is just amazing how quickly the Chinese and other adversaries are advancing,” Mr. Meink said. “Four or five years ago, we had been dominating in launch, and you can see them making massive improvements, really trying to catch up.”
As for Booz Allen Hamilton, Mr. Rozanski said it’s focused on the space domain. He emphasized that the company is the first to put an AI large language model in the International Space Station.
“This is the kind of work that we are doing,” he said. “We’re deeply committed to making sure that our country is strong.
“It is all about technology.”
• Vaughn Cockayne can be reached at vcockayne@washingtontimes.com.
• Guy Taylor can be reached at gtaylor@washingtontimes.com.
Please read our comment policy before commenting.