DECRYPTION – While the main players in AI are strengthening their ties with the Pentagon, Anthropic is speaking out in dissonance by refusing to grant the American government a free hand on the military use of its Claude model. The company is trying to set ethical limits to a technology with abysmal capabilities.
No sooner had Anthropic severed its ties with the Pentagon over ethical differences than its competitors rushed into the breach. OpenAI took the plunge in early March, followed by Google a month later. Their generative artificial intelligence models will be gradually integrated by the American Department of Defense for classified operations. xAI is not to be outdone…; the AI start-up created by Elon Musk would also have taken advantage of the break with Anthropic to reach an agreement with the Pentagon.
From then on, a dividing line is emerging in Silicon Valley. There are those like Peter Thiel or Alex Karp, respectively co-founder and president of Palantir, who take a maximalist approach to the integration of artificial intelligence in the defense and security industry. The company’s manifesto is not nuanced: “The question is not whether AI weapons will be built ; it’s knowing who will build them…
This article is reserved for subscribers. You still have 88% to discover.
Flash Sale
-70% on your subscription. No commitment.
Already subscribed? Log in
/2025/09/05/daic-template-live-68baadad16e59550656736.png "DIRECT. War in the Middle East: Iran threatens \"heavy response\" after the attack on two of its oil tankers by the United States")
