Rob Mulla

The Automation Advantage in AI Red Teaming This paper analyzes Large Language Model (LLM) security vulnerabilities based on data from Crucible, encompassing 214,271 attack attempts by 1,674 users across 30 LLM challenges. Our findings reveal a...

Recently, I’ve been cooking up some analysis of AI red teaming attacks, sourced from Crucible challenges at Dreadnode. Excited to share our new paper, focusing on how automation can amplify offensive AI techniques:

arxiv.org/abs/2504.19855

What's your take on the growing dominance of automated attacks and the implications for AI red teams? Here's ours— based on our analysis of 30 LLM challenges, attempted by 1,674 unique Crucible users, across 214,271 attack attempts: arxiv.org/abs/2504.19855

Huge news today! Excited for the future.

NEW Crucible Challenge: DeepTweak, an exploration of reasoning model behavior. Cause enough confusion 😵‍💫, retrieve the flag.

Think fast; The first three users to solve DeepTweak will be announced Friday!

➡️ https://crucible.dreadnode.io/challenges/deeptweak?utm_source=social&utm_medium=social&u…

A picture of two super soft, charcoal-colored Dreadnode t-shirts

Can't wait to see all your beautiful faces at #shmoocon

Catch @rad-ads.bsky.social , @robmulla.bsky.social and me for plush @dreadnode.bsky.social swag. I'll be at lobbycon.

*still on the hunt for a ticket 🙏

Been working on @dreadnode.bsky.social's Crucible AI CTF and just completed the "What's the flag #6" challenge. Such a fun time! Everyone in chat had a great time providing suggestions.

Hats off to the CTF authors, they did a fantastic job!

www.youtube.com/live/YTZft0L...

"Steady effort pays off, even if not always in an immediate, tangible way." - Garry Kasparov

A clean `sudo apt update` && `sudo apt upgrade` after it's been a while makes me feel warm and fuzzy inside.

That was me! I'm the Rob.

Happy Thanksgiving folks! 🦃

Spending a month on Bluesky has created a whole new model of a fight we need to be having right now. Not liberal vs. conservative or whatever, but respect vs manipulation.

I am tired of being manipulated everywhere I go. That’s why I like it here, I feel like I’m in charge for once.

This map has nothing to do with elections, and everything to do with dataframes.

R vs Pandas! Which is the better dataframe ecosystem?

Source: US google search trends over the past 5 years.

#DataScience

Let’s see what this thing is about.