2026-05-01 16:47
publication
2026-05-01 16:47
The Pentagon announced on Friday that it had concluded agreements with seven companies from the artificial intelligence (AI) sector regarding the use of their tools in the secret work of the US Ministry of Defense.
The ministry announced the conclusion of contracts with the following companies: SpaceX (owned by xAI), OpenAI, Google, NVIDIA, Reflection, Microsoft and Amazon Web Services. Some of them already cooperate with the Pentagon, but the agreements announced on Friday allow the ministry to use their products more widely. According to the ministry, the companies agreed to allow the Pentagon to use their technology “for any lawful purpose.”
Anthropic loses contracts because it does not want AI in weapons systems
The list does not include Anthropic, with which the Pentagon is in a dispute over security measures regarding how the military uses the company's AI tools, Reuters noted. The company has requested that its technologies not be used for autonomous weapons and mass surveillance.
As CNN reported, until recently, Anthropic's Claude was the only AI model available in the secret networks of the US Ministry of Defense, i.e. highly secured systems used, for example, to plan military missions.
The Pentagon formally recognized this year that Anthropic poses a “supply chain threat” by prohibiting defense contractors from doing business with the company. This is the first time this classification – previously reserved for companies from China and other U.S. adversaries – was used against an American entity.
“This is not a theoretical threat.” Revolt at Google against the Trump administration
Google does not want to be used by the American administration, and they are particularly afraid of putting artificial intelligence in the hands of Trump's associates. – We want AI to serve humanity and not be used in an inhumane or extremely harmful way – they wrote in a letter quoted by the Financial Times. This is especially true when it comes to secret military operations.
Cloud reinforcements. The US Ministry of Defense is expanding cooperation
The US Ministry of Defense and the White House emphasize that US law and government institutions, and not the internal regulations of a private company, must decide how to use artificial intelligence systems for military purposes.
Anthropic points out that even the best AI models are not reliable enough to be part of autonomous weapons and using them for this purpose would be dangerous.
CNN emphasized that the so-called one big beautiful bill with President Donald Trump's assumptions provides for a large sum of money for the Pentagon for the development of artificial intelligence and offensive cyber operations. Tech companies are fighting for this money.
The Pentagon also announced on Friday in a statement that the ministry's main AI platform, GenAI.mil, is used by 1.3 million ministry employees. It has been in use for five months.
From Washington Natalia Dziurdzińska (PAP)
Sun/Apr/
