The United States military reportedly used advanced artificial intelligence to help identify and strike roughly 1,000 targets within the first 24 hours of its campaign against Iran and at the centre of that operation was technology built by Palantir and Anthropic, according to a report by The Washington Post. The system in question is called the Maven Smart System, developed by data-mining company Palantir, which ingests classified intelligence from satellites, surveillance platforms and a range of other sources. Embedded within it is Claude, Anthropic's generative AI model. The scale of what's being described here is significant. Here's what we know.

What Claude Was Actually Doing Inside Maven

According to The Washington Post, which cited three people familiar with the system, Claude was integrated into Maven to analyse intelligence data, suggest potential strike targets and rank them based on operational importance. Two people familiar with the system told the newspaper that Maven, powered by Claude, generated hundreds of target suggestions, produced location coordinates and helped US planners prioritise during the Iran campaign. One person told The Washington Post that the combined system has fundamentally changed the pace of military planning, converting processes that previously took weeks into near real-time operations. The AI tools were also reportedly used to assess strike outcomes after they were carried out.

A First for Large-Scale Combat Use

While Claude had previously been deployed in security contexts including counterterrorism operations and, according to two people cited by The Washington Post, the raid that led to the capture of Venezuelan President Nicolás Maduro, the Iran campaign represents its first known use in large-scale military combat operations. Over the past year, military planners have reportedly expanded the system's use across multiple branches of the armed forces, with more than 20,000 military personnel using it as of May last year. Rear Admiral Liam Hulin, deputy director of operations at US Central Command, noted in a 2024 talk that the system draws on intelligence from 179 different data sources.

Operation Epic Fury Update: Inside CENTCOM's List Of US Assets Used, Targets Hit Amid Israel-Iran Tensions

The Ban That Came Hours Before the Bombs

The timing of what happened next is striking. Hours before the Iran bombing campaign began, US President Donald Trump announced a ban on Anthropic's AI tools across government agencies, according to The Washington Post. The administration has given agencies six months to phase out the technology, following disputes over its use, particularly around mass domestic surveillance and fully autonomous weapons systems. Despite the ban, two people familiar with the matter told the newspaper that the military will continue using the technology during the transition period while a replacement is developed. One person was blunt about the Pentagon's position: "We're not going to let [Anthropic CEO Dario Amodei's] decision making cost a single American life."

Anthropic CEO Dario Amodei, for his part, wrote in a blog post during negotiations with the Pentagon that he believes deeply in "the existential importance of using AI to defend the United States and other democracies." The company was among the first major AI firms to work with classified US government data.

Rivals Moving In, Experts Raising Concerns

Source: India Latest News, Breaking News Today, Top News Headlines | Times Now