There is an important facet to Anthropic's report on malicious AI use: it shows the extraordinary risk of using frontier AI for state-sponsored covert ops and secret weapons programs. You may be able to go faster and cheaper, but now everyone can see you.