Anthropic CEO says AI company ‘cannot in good conscience accede’ to Pentagon’s demands
Summary
AP reports that Anthropic CEO Dario Amodei says the company cannot in good conscience accede to Pentagon demands to broaden use of its Claude AI. Anthropic warns that new contract language could enable mass surveillance or fully autonomous weapons, while the Defense Department argues it seeks lawful, mission-ready use and warns of potential remedies if terms aren’t accepted. The story highlights tensions between AI governance, national security, and private-sector risk appetite, and it intersects with ongoing calls in Congress for stronger AI governance.