Anthropic vs. Pentagon: AI Ethics & Autonomous Weapons

Tense emails show a disagreement between AI lab Anthropic and the Department of Defense. Court documents reveal the exchange that led to the two organizations falling out earlier this year. Anthropic CEO Dario Amodei was concerned the Pentagon would use his company's AI for purposes he found unacceptable. Specifically, he worried about its integration into autonomous weapons systems and domestic surveillance tools. The Pentagon, however, maintained it could use the technology for "all lawful uses." The dispute began in January when Emil Michael, a Pentagon official, reached out to Amodei. Michael hoped Anthropic would agree to the Pentagon's demands. Amodei responded by insisting on guardrails, including banning the use of AI for fully autonomous weapons and domestic surveillance. Michael called this "not workable." He warned that there was one more chance to align before parting ways. He also stated that "there is no distinction in our world between weapons that are defensive or offensive." Amodei pointed out that "all lawful uses" was problematic for Anthropic because US law permits domestic surveillance. He noted that the Pentagon seemed to "completely remove our redlines." The next day, Defense Secretary Pete Hegseth designated Anthropic as a supply-chain risk, ending negotiations. This situation highlights the ongoing debate about the ethical use of advanced AI technology.