The company’s Claude chatbot is one of the few AI systems cleared for use in classified settings. But a standoff between Anthropic and the Trump administration is putting its government work at risk.

(Image credit: STAFF/AFP via Getty Images)