Anthropic says it will not remove AI safeguards despite Defense Department pressure
Feb 27th 2026
CEO Dario Amodei says Anthropic has deployed Claude across classified US systems but will not drop safeguards for mass domestic surveillance or fully autonomous weapons despite reported Defense Department demands and threats.
- Anthropic says its Claude models are widely deployed in Defense Department and other national security systems for mission critical tasks.
- CEO Dario Amodei says the company was first to deploy models in classified networks, national laboratories, and to provide custom national security models.
- Anthropic says it blocked access by firms linked to the Chinese Communist Party, stopped CCP sponsored abuse attempts, and supported tighter chip export controls.
- The company refuses to allow its models to be used for mass domestic surveillance because of privacy and democratic risks.
- Anthropic also will not knowingly provide technology for fully autonomous weapons, citing current reliability limits and a lack of adequate guardrails.
- Anthropic says the Defense Department demanded companies accept "any lawful use" and threatened removal, a supply chain risk designation, and use of the Defense Production Act; the company says it will not acquiesce and will help transition services if offboarded.