AI Firms Resist Pentagon's Push for Unrestricted Military Use of Models

The US Department of Defense is in talks with several major artificial intelligence firms, including Anthropic, OpenAI, and xAI, over the terms under which they can be used to support military operations. At stake are issues of autonomous lethal weapons, mass surveillance, and other sensitive applications that raise significant concerns about accountability, transparency, and the potential for human rights abuses. In a bid to prevent the deployment of autonomous killing machines without adequate safeguards, Anthropic CEO Dario Amodei has been leading the charge against Pentagon demands for greater access to its AI models. His company has thus far refused to agree to new contract terms that would allow the military to use its technology without strict oversight or limits on its application. Industry insiders describe the negotiations as a classic case of a ‘supply chain risk’ – with the Pentagon seeking to label Anthropic and other companies with high-priority clearance, effectively giving them access to sensitive government contracts and resources. However, Anthropic’s stance remains unwavering, driven by concerns about the long-term implications of military exploitation of AI technology. The stakes are significant: if the Pentagon succeeds in securing unrestricted access to these powerful tools, the US could potentially find itself at the forefront of a new era of autonomous warfare, with devastating consequences for human life and global stability.