Pentagon threatens to cut off AI contract over ethical standoff
The US Defense Secretary just told Anthropic: let the military use your Claude AI without restrictions, or risk losing your contract by this Friday.
Right now, Anthropic blocks things like mass surveillance of Americans and fully autonomous weapons—rules the Pentagon says are holding them back.
Outcome could set new rules for defense tech companies
This standoff could shake up who gets to build AI for the military—and how ethically they do it.
If Anthropic doesn't budge, it could lose out on $200 million in future deals, while rivals differ: xAI has reportedly agreed to 'all lawful use' at any classification level, and OpenAI has shown flexibility on unclassified work and continues negotiating classified access terms.
The outcome might set new rules for how far defense tech companies can go with their AI, especially when balancing power and responsibility.