Vice President JD Vance said he is concerned about how artificial intelligence will be used in warfare and urged the military on Thursday to be cautious about the technology, remarks he made as the Pentagon moves forward with using AI on the battlefield.
Speaking to graduating cadets at the U.S. Air Force Academy in Colorado Springs, the vice president — who served in the Iraq War during his time in the Marines — acknowledged the growing unease among Americans as AI stands to “affect the labor market” and has “fundamentally changed how we interact with one another.”
“But the thing I worry about most with AI is how it will change warfare,” Vance said, adding that it “already has.”
“If the warfare of the future is to live up to the moral values of our ancestors, decisions over life and death must be made by humans and not machines,” he said.
The Pentagon has moved quickly to integrate AI into its war-fighting, even as it faces questions about how to impose appropriate safeguards on the still-new technology.
The vice president, a Catholic who has previously been bullish about the adoption of AI, used the Air Force Academy speech to “endorse” a sentiment expressed by Pope Leo XIV in his encyclical on AI released this week, even as some other Trump administration officials have dismissed Leo’s new guidance.
Vance praised the pope’s instruction “not to outsource the most important moral decisions to digital technology,” saying he wanted the cadets to take it to heart as it applies to decisions made in war.
“So as AI transforms the battlefield — in some ways positively, in some ways not — I ask that you be jealous and selfish about your role as a decision-maker in warfare,” he said. “Use technology to make you better, but never submit to it. You are the masters of warfare, and both your minds but also your hearts are the opposite of artificial.”
Vance’s broad message about the importance of human judgment comes as the Trump administration struggles to articulate more practical guidelines for the use of AI in the United States. Last week, President Donald Trump decided at the last minute not to sign a long-awaited executive order on AI after several tech executives urged him to hold off, concerned it would restrict the technology’s development.
Other administration officials have brushed off warnings about embracing AI too quickly and without proper evaluation. In January, Defense Secretary Pete Hegseth said the military was in “a race” to develop AI-enabled warfare.
“We must accept that the risks of not moving fast enough outweigh the risks of imperfect alignment,” Hegseth wrote in a January memo directing broad new implementation of AI development and integration across the military. “Speed wins.”
When faced with the indirect criticism from Leo in recent months about the administration’s involvement in the Iran war, particularly that it did not meet the standards of a “just war,” Hegseth downplayed the significance of the pope’s opinion on military matters.
“The pope’s going to do his thing, that’s fine,” Hegseth told reporters last month. “We know what our mission is, we know what authority we have. We’re very clear about that, we follow that, the orders of the president.”
The Pentagon is still under scrutiny for a military strike on a school in Iran in the first hours of the war. It killed at least 175 people, many of them schoolgirls, according to Iranian officials. The school was hit as the U.S. struck a blistering 1,000 targets in those first 24 hours, enabled in part by the most advanced AI the Pentagon has ever used in warfare, the Maven Smart System.
That system used the AI tool Claude to suggest hundreds of targets, issue location coordinates and prioritize targets according to importance.
While the investigation into the Iranian school strike is ongoing, officials who have previously spoken to The Washington Post said it is likely the system was fed erroneous targeting information. The school, which is adjacent to an Iranian military site, had been cleared as a legitimate target and had been identified as a weapons factory or storage building.
The Pentagon had previously used the Claude-enabled Maven system in January during its raid to capture Venezuelan President Nicolás Maduro. In response, the company that makes Claude, Anthropic, raised questions as to how the Pentagon was using its AI tool, given the firm’s stated ethical restrictions.
The debate boiled over and resulted in Trump banning Anthropic from future government contracts, while Hegseth claimed Anthropic was a supply chain risk.
The matter is still in litigation, and meanwhile, the U.S. military used Anthropic technology as it went into Iran.
Dan Lamothe contributed to this report.
The post Vance says military should never let AI make life-and-death decisions appeared first on Washington Post.
