Overview
- OpenAI amended its new Defense Department contract to state its models shall not be intentionally used for domestic surveillance of U.S. persons, with any intelligence‑agency use such as by the NSA requiring a separate modification.
- CEO Sam Altman acknowledged the rollout was rushed and said the company clarified terms after employee and public backlash, which included uninstall campaigns and critical app‑store reviews.
- The Pentagon designated Anthropic a supply‑chain risk and President Trump ordered a six‑month phase‑out of Claude across federal agencies, prompting contractors and defense tech startups to begin replacing Anthropic’s tools, with Reuters reporting Lockheed Martin is expected to remove them.
- Anthropic said it will challenge any formal designation in court and reiterated it will support U.S. national security under its stated red lines against mass domestic surveillance and fully autonomous weapons.
- Multiple outlets reported that Claude supported recent U.S. strikes on Iran, which neither the Pentagon nor Anthropic confirmed, intensifying scrutiny of how deeply military workflows rely on commercial AI and underscoring calls for Congressional rules beyond contract terms.