Follow Al Rai
A US appeals court on Friday supported the Pentagon's decision to exclude Anthropic from its supply chain, a decision made when the AI company refused to remove certain safeguards from its 'Claude' model.
The dispute began in February when Anthropic rejected Defense Secretary Pete Hegseth's request to remove restrictions on the use of the 'Claude' model in autonomous lethal weapons and mass surveillance operations of Americans within the country.
The Pentagon's classification of the company in March as a risk to the supply chain led to the cancellation of military contracts with Anthropic and prevented other contractors with the US Department of Defense from using its technology.
The Pentagon immediately began collaborating with the company's competitors.
Agreements were signed with Open AI, the creator of innovative Chat GPT, on the same day the ban on Anthropic was imposed, with Elon Musk's X AI, the developer of the Grok model, in February, and with Google and Microsoft a few months later.
Since then, Anthropic has been shunned by the White House, despite playing a major role in the AI sector and heading towards a massive initial public offering expected in weeks with a valuation close to a trillion dollars.
The White House also expressed dissatisfaction with the repeated warnings issued by Anthropic CEO Dario Amodei about the risks of artificial intelligence, which US President Donald Trump described as a 'hoax'.
US Assistant Secretary of Defense Emile Michael celebrated the ruling in a post on the X platform, saying, 'The hammer of justice has crushed Anthropic's arguments'.
He added, 'Warriors will perform better knowing that no private company will interfere with the military chain of command'.
This appeals court decision contradicts a separate ruling by a federal court in California last month that the Trump administration's ban on Anthropic was unlawful.
The appeals court on Friday supported the Pentagon's decision based on public procurement laws.
An Anthropic spokesperson said in a statement to US media, 'We respectfully disagree with the court's decision. Another federal court has already ruled the government's parallel classification to be unlawful. We are confident in our position and are exploring all available options, including requesting additional review'.
The US military has used the 'Claude' model in a range of sensitive and classified systems.
According to court documents, this included the operation in January that resulted in the arrest of Venezuelan President Nicolas Maduro.
Know more about this story?Send us a tip
Ask about this report — answers stay grounded in its text.
