Claude Code Is Quietly Fingerprinting China-Linked API Routers

https://www.vincentschmalbach.com/claude-code-china-router-fingerprint/

by vincent_s • 23 days ago

9 4 23 days ago

This post is full of Claude's irritating verbal ticks ("It's not X, it's Y"). It is fine to use Claude (or another LLM) to help with this kind of reverse engineering, but I would prefer the write-up be done by hand.

d-lo• 23 days ago

There's another good write up and discussion on Reddit (with fewer LLM writing ticks): https://old.reddit.com/r/ClaudeCode/comments/1ujilqt/anthrop...

I do wonder if Anthropic is modifying output via prompt modification or some of the Fable style weights adjustments for requests that contain these sentinel values. That would be one way to try to prevent distillation, and they have shown a willingness to silently modify model behavior for user input they deem dangerous.

Chu4eeno• 23 days ago

They have always sent a ton of analytics + everything you do, is this news just because the Chinese are afraid they'll get cut off?

samuelknight• 23 days ago

The Reddit poster called it spyware which is not right. They are using stenography to fingerprint accounts using Claude Code instances routing through Chinese resellers.

Claude Code Is Quietly Fingerprinting China-Linked API Routers

Related Stories