Sheera Frenkel
As the war in the Middle East continues, intelligence gathered by the Pentagon is being analysed by technology from the artificial intelligence company Anthropic, on a system run by the data analytics firm Palantir.
Drones created by a defence tech start-up in Arizona have emerged as a key piece of the US war arsenal. And anti-drone systems made by a California start-up have been deployed to protect American forces in the region.
Silicon Valley made risky bets in recent years on developing defence-related technology and providing services to the US military establishment. Now those bets are paying off. From behemoths providing data systems to smaller companies offering novel weapons, tech firms such as Google, Palantir and OpenAI have found themselves at the heart of the US war effort.
Their central role amounts to an “I told you so” moment. For years, the tech industry’s efforts on defence-related offerings faced scepticism and opposition, with no clear or immediate business rewards. Many Silicon Valley engineers opposed the use of powerful technologies for killing, battles and other military purposes, and those concerns persist.
Despite those fears, venture capital firms have poured billions of dollars since the 2010s into start-ups building drones, lasers and other military systems. In January, Andreessen Horowitz, the firm founded by entrepreneurs Marc Andreessen and Ben Horowitz, closed a new, almost $US1.2 billion ($1.7 billion) fund to invest in defence technologies.
In recent years, defence tech start-ups often ploughed ahead with weapons prototypes before they had official government contracts. At the same time, executives including Palantir chief executive Alex Karp started cultivating more ties with the government.
As president, Joe Biden welcomed military technology, and Donald Trump has further embraced it. Last year, Trump issued an executive order calling for the military to update its system for acquiring technology so it could incorporate new tools faster. His domestic policy bill last year allocated $US1 trillion to defence in 2026, including for technology offered by defence tech firms.
Now the war has cemented that work, most likely leading to more business between the tech industry and the military.
This month, OpenAI chief executive Sam Altman agreed to bring his company’s AI systems onto the Pentagon’s classified networks. Google signed a deal to bring AI bots known as “agents” into the US Defence Department. On Friday, the US Army said it had awarded Anduril, a defence tech company, a $US20 billion deal for AI-backed software to run on military systems.
“People are pointing to this moment as a proof point,” said Garrett Smith, a former lieutenant colonel and the CEO of Reveal Technologies, which makes mapping technology for the army.
‘I wouldn’t be surprised if this is called America’s first AI war.’
Retired Air Force lieutenant general Jack Shanahan
“It has shown us that in creating and selling these technologies to the US military, we are on the right track. We have made the right investments.”
Pentagon officials said they were excited about how well new technology like AI-related systems had performed in the US-Israeli war against Iran. Two officers who were not authorised to speak publicly said the war was an inflection point in showing how modern technology could work with existing military systems.
Jack Shanahan, a retired air force lieutenant general who led the Pentagon’s efforts to incorporate AI during the Biden administration, said the war would speed the adoption of more technologies. Over the next decade, he expects legacy and modern systems to be melded together, along with more powerful AI.
“I wouldn’t be surprised if this is called America’s first AI war,” Shanahan said.
But Amos Toh, a senior counsellor at the Brennan Center for Justice, a New York non-profit focused on law and public policy, cautioned that this gung-ho attitude might lead to little oversight of new systems and an overreliance on just a few tech companies. The military and the government need “to take a look at the dependencies it is creating”, Toh said.
OpenAI, Google and the Pentagon did not respond to requests for comment. (The New York Times has sued OpenAI and Microsoft, accusing them of copyright infringement of news content related to AI systems. The two companies have denied those claims.)
Project Maven, an AI-backed system built by Palantir for the Pentagon, is one prominent example of modern technology in the war. Maven works by layering Palantir’s data systems with AI technology from Anthropic. Anthropic’s systems analyse real-time data about battles and other war scenarios, while Palantir’s technology draws conclusions about which targets to strike.
Airstrikes hit more than 2000 targets in Iran in the first four days of the war. Many of the targets were selected from a list produced by Maven after it analysed information from drones, satellite imagery and other sources.
On Thursday, Palantir’s Karp said in a CNBC interview that AI was giving US forces an edge. Palantir’s stock has soared more than 12 per cent since the war began.
“Our adversaries and enemies are witnessing our ability to fight that they don’t have, and they will find it very hard to acquire,” he said. “America is the centre of the AI revolution.” Palantir and Anthropic declined to comment.
Technologies from defence start-ups are also being deployed. A system that uses drones to counter other drones, called Merops, developed as a venture project by former Google CEO Eric Schmidt, has become instrumental to protecting US assets in the conflict. The system, which is small enough to be launched from the back of a pick-up truck, uses AI to seek out and intercept drones before they can reach their targets.
Schmidt’s office declined to comment.
Small and lightweight drones called LUCAS (low-cost uncrewed combat attack system) from SpektreWorks, a start-up in Phoenix, have also been deployed in the battlefield. The LUCAS drones, which mimic Iran’s Shahed drones, are designed for one-way flights. They have been effective in overwhelming missile defence shields and engaging in the type of drone warfare first made popular in Ukraine, a US official said.
SpektreWorks declined to comment on its work with the Pentagon.
This month, US Central Command posted footage of rows of the drones as they were readied to be sent to US forces.
“I’d like to point out these drones were originally an Iranian design,” said Admiral Brad Cooper, the head of the Central Command. “We took them right back to America, made them better and fired them right back.”
This article originally appeared in The New York Times.
The Market Recap newsletter is a wrap of the day’s trading. Get it each weekday afternoon.
