2/ Every AI router on the market can currently read the prompts passing through it. That's just how it works: your request has to be decrypted somewhere before it reaches the model, and that somewhere is the router's own servers.
"We don't log your prompts" is only a policy.
Login to reply
Replies (1)
3/ And the router in the middle is exactly where things go wrong. In April of this year:

X (formerly Twitter)
Chaofan Shou (@shoucccc) on X
26 LLM routers are secretly injecting malicious tool calls and stealing creds. One drained our client $500k wallet.
We also managed to poison rout...