Researchers Uncover 'LLMjacking' Scheme Targeting Cloud-Hosted AI Models

Researchers have uncovered a new attack called "LLMjacking" that targets large language models (LLMs) hosted on cloud services.

1 comments

This has a similar feel to those cheap "international call cards" that uses hacked PBX systems through their own proxy so their users don't know it's going through random hacked systems