Rabbit on Nostr: In my last response against an AI assisted attacker who knew what they were doing, it ...

In my last response against an AI assisted attacker who knew what they were doing, it was like playing WarGames - by the time I identified a compromised ID, there was nothing left but a smoldering hole- everything that identity had access to had been exploited and lateralled in moments. Effectively instantly.

Even a large team of seasoned red teamers can't enumerate access paths and pivot that fast. Even with scripts (unless the scripts are basically doing what the LLM does)