Episode 16: The Hacker's Toolkit
Critical Thinking - Bug Bounty Podcast
00:00
New Attack Vectors for AI Integrations With Web to Technology
Rezzo is talking about prompt injection and some indirect prompt injection and then polyglot prompt injection. These are all LLM based injection attacks, sort of new attack classes or attack vectors designed to exploit sort of AI integrations with web to technology. Plug-in hijacking, I think is a really cool one as well. So definitely looking forward to seeing these attacks around.
Transcript
Play full episode
