This blog has been quiet since 2014, but I can see people still come here, mostly for the Hydra and Linux privilege escalation posts. Thanks for reading. Those posts are old now, so check the tools and versions before you rely on them.
These days I write at ShortSpan.ai. It covers AI security research: prompt injection, jailbreaks, agents that can be talked into things, and what all of that means for people who test systems for a living. Each piece is a short review of one paper, written for practitioners rather than academics.
A few recent ones, if you want a feel for it:
- An open-source pentest agent goes to work in the wild
- Obfuscated jailbreaks get past every model tested
- Prompt injection defences miss attacks that ask for less
There's a weekly newsletter on the site if you'd rather it came to you.
One more thing. The insidetrust.com domain is no longer mine. Anything at that address isn't from me. This blog and ShortSpan.ai are the places to find me now.
Ben
