In graphic detail: How Anthropic’s Pentagon refusal is paying off in downloads, brand trust and enterprise deals

Anthropic’s war with Washington last week is winning in the public.

The $200 million it walked away from by refusing the Pentagon’s demands may turn out to be the best marketing spend in Silicon Valley for years. Its principled stand has driven a surge in users and supercharged its brand — no small feat for a company that has always bet on enterprise over the masses. What makes this all the more remarkable is that it happened in the same week Anthropic quietly loosened its own safety commitments, dropping a hard pledge to pause training more powerful models if it couldn’t guarantee were safe. The public, it seems, is responding to the fight it can see — not the fine print it can’t.

And in doing so a moral line is being drawn — real or perceived — between Anthropic and its rivals. OpenAI, which stepped in to take the Pentagon contract Anthropic refused, now finds itself on the wrong side of it.