Appeals court allows Pentagon to label Anthropic a national security risk (www.businessinsider.com)

A federal appeals court has ruled that the Pentagon can blacklist AI company Anthropic from all defense contracts, determining that its AI products pose a national security risk due to embedded code that restricts certain functionalities. The D.C. Circuit Court of Appeals concluded that the Defense Department had adequate grounds to perceive Anthropic’s model, Claude, as a potential supply chain risk, given the company's self-imposed limitations on usage, including prohibitions against autonomous weapons or mass surveillance applications. This ruling significantly empowers the executive branch to label both domestic and foreign companies as security risks, promoting a broader authority to protect national interests in technology. Anthropic has expressed concern that this designation will inflict considerable reputational damage and lead to substantial financial losses, potentially reaching hundreds of millions in revenue. The implications extend beyond Anthropic, signaling a more aggressive stance on AI governance and the scrutiny of tech companies engaged in defense-related contracts, shaping the landscape of AI development within national security frameworks.