John Ratcliffe, the Director of the U.S. Central Intelligence Agency (CIA), declared on Tuesday that the risks posed by cutting-edge artificial intelligence (AI) models are comparable to digital nuclear weapons.
According to reports from AFP and The New York Times, Ratcliffe made this statement during his address at the Amazon Web Services (AWS) Summit in Washington, D.C. He revealed that he has been engaging in discussions about the impact of frontier AI models with numerous presidential advisors on national and economic security.
Ratcliffe emphasized that likening these AI capabilities to digital nuclear weapons is not an exaggeration, suggesting that major AI companies are developing models that could be seen as doomsday devices.
He assessed that the rapid advancement of emerging technologies, particularly AI, is reshaping global geopolitics. Ratcliffe warned that U.S. adversaries are attempting to steal and manipulate American technological progress for their own agendas, noting that AI development is significantly raising the stakes in international competition.
Ratcliffe disclosed that the CIA has streamlined its technology procurement processes to accelerate the adoption of new technologies. He also mentioned that the agency has undergone a major organizational overhaul centered on cyber operations.
The CIA director stated that the agency will take calculated risks, conduct experiments, and adapt its approach as needed. He stressed that more agents must become as proficient with lines of code as they are with human sources.
However, Ratcliffe emphasized that despite embracing AI, human decision-making will continue to drive the CIA’s operations. He asserted that only humans can determine the right course of action.
Experts note that it’s highly unusual for the director of a key U.S. intelligence agency like the CIA to make such public statements.
Recently, the Donald Trump administration has recognized high-performance AI models as a national security risk on par with nuclear weapons and has been expanding efforts to regulate them. Several policy think tanks describe this as the U.S. entering a technological arms race with competitors like China.
On June 12, citing national security concerns, the U.S. government ordered AI company Anthropic to block all foreign nationals from accessing its top-tier models, Fable5 and Mythos5. The export restrictions were partially eased for Mythos on June 26 and will be fully lifted on July 1.
OpenAI, the developer of ChatGPT, also unveiled its cutting-edge model GPT-5.6 but decided to limit its availability to select partners at the request of the U.S. government.