Statement from Dario Amodei on Our Discussions with the Department of War (www.anthropic.com)

Dario Amodei, CEO of Anthropic, has issued a statement addressing the company’s collaboration with the Department of War, emphasizing the pivotal role of AI in defending democracies against autocratic threats. Anthropic has become a leader in integrating its AI models, including Claude, into the U.S. government’s classified networks, focusing on mission-critical applications such as intelligence analysis and operational planning. The company has taken significant steps to ensure that its technology does not benefit adversaries, notably refusing to engage with entities linked to the Chinese Communist Party, highlighting a commitment to national security even at significant financial cost. However, Amodei underscored two critical areas—mass domestic surveillance and fully autonomous weapons—where Anthropic refuses to compromise, as they pose risks to democratic values and public safety. Currently, the Department of War is seeking to contract with AI firms that would forgo such safeguards, raising concerns about the ethical implications of deploying advanced AI technologies. Amodei expressed hope for continued collaboration with the Department under secure conditions while preparing for a potentially abrupt transition if their stance doesn’t change. This discussion reveals the ethical tensions present in military applications of AI, as companies balance innovation, national security, and democratic principles.