Last week, a federal court upheld the Trump administration’s designation of Anthropic as a “supply chain risk,” a scarlet letter that would effectively label the company a national security threat. It was a major win for the federal government in its long-running dispute with the AI startup, which earlier this year said it would not allow the Pentagon to use its technology for domestic surveillance or autonomous weapons systems.
The legal battle—which has yet to be settled in court—comes down to a single question: Who should have the final say over how American-made technology gets used by the military? And by extension, who has the power to shape American national security interests? Historically, that power rested solely and unambiguously in the hands of the federal government. Today things are much murkier—and much more dangerous, according to the author and longtime national defense journalist Sharon Weinberger, author of the new book, Valley of Death. “This idea that a private company should dictate to the Pentagon how technology is used… this is what you’re seeing now,” she told Gizmodo in an interview.
Military technology built by traditional defense contractors was often designed in collaboration with and ultimately owned by the U.S. government; the Pentagon never had to get Lockheed Martin’s permission before using the F-22 to conduct bombing raids in Afghanistan. That dynamic has shifted in recent decades throughout the tech boom in Silicon Valley, which has given rise to the wealthiest companies the world has ever known. The Air Force today is largely dependent on rockets made by Elon Musk’s SpaceX; satellites built by Starlink—another Musk-owned company—have played a pivotal role in the Ukrainian military’s fight against Russia; and a host of other privately backed startups have been competing with traditional defense firms for lucrative Pentagon deals to build software and actual, physical weapons for the U.S. government as it seeks to rev up defense spending to prepare for a potential military conflict with China.
President Dwight Eisenhower—a former five-star general and the Supreme Commander of Allied forces in Europe during the Second World War—warned Americans during his 1961 farewell address about the danger of allowing national security to be controlled by the interests of private companies. “In the councils of government, we must guard against the acquisition of unwarranted influence, whether sought or unsought, by the military-industrial complex,” he said, coining a now-famous phrase. “The potential for the disastrous rise of misplaced power exists and will persist.”
We’re now living through the rise of a “Pentagon-Silicon Complex,” Weinberger argues in her book. Unlike the military-industrial complex of old, she writes, the new paradigm between the defense industry and the private sector is increasingly dictated by the whims of mercurial tech CEOs, many of whom see their companies—and themselves—in distinctly messianic hues.
Musk, Palantir’s Alex Karp, Anduril’s Palmer Luckey, and an armada of lesser-known but no less ambitious other tech executives brazenly style themselves defenders of Western democracy and the vanguard of a new era for American prosperity. These executives “have opinions—not just about technology, but also about politics, the border, immigration, America’s enemies, birth rates, the way we should be waging war, and crucially, whom we should be waging war against,” Weinberger writes. “These convictions… are what makes their rise to power so consequential—and so urgent.”
That rise, she argues, has been fueled largely by vibes: portraying a sci-fi vision for the future of warfare that’s thus far mostly untested in actual combat settings, but has nonetheless proven to have significant appeal among investors. Palantir, for example, has gone to great lengths to position itself as an AI company, even though it doesn’t train its own AI models. It also pushes a distinct vision for the future of warfare, one where almost everything is autonomous, and the outcomes of battles are decided by the press of a button. In a 2024 commercial that Palantir aired during the Army-Navy football game, a military commander is shown donning a pair of futuristic smart glasses and clicking “Confirm” on a digital screen, immediately launching a fleet of drones that attack a fleet of enemy navy ships. Then a message appears: “Battles are won before they begin.” (Weinberger notes in her book that “the type of autonomous technologies featured in the commercial have not yet been fielded, and Palantir isn’t building them.”)
Palantir’s valuation, she told me—which at the time of this writing stands close to $455 billion—a roughly 500% increase compared to the same time just two years ago—”is not tied to its revenue, which is a fraction of what Lockheed Martin, RTX [formerly Raytheon Technologies], and Northrop Grumman earned. It’s tied to their narrative… [and] that’s done very well for them. It’s working.”
Anduril—a company named after a sword in the Lord of the Rings whose full title is Anduril: Flame of the West—has marketed itself as a defender of democratic values around the world. In an X post last summer, Luckey claimed his company “had people/hardware in Ukraine two weeks into the invasion,” and that its “autonomous weapons have destroyed hundreds of millions worth of Russia’s war machine.” But Luckey’s heroic rhetoric soon ran into some practical hurdles: The Wall Street Journal reported in November that many of the drones Anduril had sent to Ukraine were prone to a number of technical glitches, causing setbacks and safety concerns. Luckey, meanwhile, was already shifting the company’s priorities to prepare for a potential conflict with China over Taiwan in 2027. War in the Pacific could be big business for Anduril, which manufactures not only drones but also autonomous cruise missiles for the U.S. military.
But war profiteering is as old as warfare itself. What is new, according to Weinberger, is the ascendance of hugely powerful defense startups that control how their products are used, and who gets to use them. “What differentiates the Palantirs, Andurils, and SpaceX’s from traditional defense companies is that they own the technology they’ve developed, placing an immense amount of power in private hands,” she writes. “That power stems not from the fact that they can sell their products to America’s enemies—U.S. law can prohibit that—but from their control over critical technology that the U.S. military needs.” Including, in theory, the most deadly technology humanity has yet invented: “I would definitely build nuclear weapons,” Luckey said in an interview with Axios earlier this year. “I would build fission weapons. I would build fusion weapons. I think that nuclear weapons have been one of the most stabilizing forces in history, ever.”
Seen in this light, the Anthropic-Pentagon dispute is less about the ethical red lines of a single company than it is a harbinger of things to come, as the military-industrial complex of yesteryear morphs into the Pentagon-Silicon Complex. Many Americans would probably applaud a private company for refusing to allow its AI systems to become Big Brother or Skynet. The peril, Weinberger argues, lies in what could happen when the whims of a powerful tech CEO don’t align with the interests of the American military or the American public.
“This is the danger of the moment we’re in,” Weinberger told me. “I’d rather have an accountable government than unaccountable CEOs.”
