Derek Banks @0xderuke

I’m generally opposed to government involvement in private companies, but watching Anthropic’s handling of Fable makes me question whether they’re the company I want arriving at AGI/ASI first. The technology is one thing. The judgment and incentives behind it are another.

@MalwareJake I am running Qwen 3.6-27B Dense with 256k KV cache and it seems pretty capable and fast so far - been using Hermes as a harness and vLLM for the inference engine. I've not had luck with Gemma4 and consistent tool calling for agentic work.

I want to share a quick thought for people in cyber security. This will be my longest tweet ever. I’ve spoken to many lately who are having an existential crisis from the constant posts about “the end of cybersecurity jobs.” Yes, things are changing quickly. This is a significant moment for the tech industry. Change can be uncomfortable. But we’ve seen cycles like this before. • When GitHub and open source took off, people said software engineers would disappear because code was free. • When AWS and cloud computing emerged, people said infrastructure jobs would vanish. • When fuzzing and SAST tools improved, people said vulnerability research would disappear. • Virtualization would eliminate infrastructure jobs. • Mobile computing was going to end desktop dev. • Exploit mitigations would end exploitability. It didn't. Each time automation improved, the amount of software grew faster than the automation. It does feel "different" this time as it's explosive. Some roles will shrink: • repetitive pentesting • basic vulnerability scanning • tier-1 SOC monitoring But other areas are expanding rapidly: • AI system security • supply chain security • identity architecture • autonomous agent security • critical infrastructure protection Historically, every time we eliminate one class of bugs, new classes emerge. Right now people are vibe-coding entire systems, giving AI access to their machines, crossing trust boundaries, and deploying autonomous agents with excessive permissions. The legal and regulatory world is nowhere close to ready. There will absolutely be new failure modes. Humans are amazing and always adapt, finding new ways to do things. The worst thing you can do right now is fall into a doom loop. ...and I’ll be honest, I too have felt the "psychological paralysis" a few times thinking, “Is this time different?” It's especially impactful when it comes from someone I respect in the community. There are certainly unknowns, in an industry where we've become accustomed to predictability. But... the majority of those reactions are usually driven by social media, not reality. Platforms like X reward engagement, and sensational doom posts spread faster than measured thinking. If you see something like: “Holy #$%^! Opus 66.6 just found every bug in Chrome and replaced 50 startups!” …mute it and move on. Instead: Stay curious. Learn the new technology. Adapt your skillsets. Build things. We’ll get through this transition the same way we always have. If I'm wrong then Sam Altman better be right about UBI! :) I'm sure that if this tweet gets any engagement that I'll get some heat for it, but a good friend of mine reminds me often to focus on what you have control over. I'll revisit this tweet at DEF CON 40!

I created a Claude and Grok integration into Signal running on my DGX Spark at home. I can have it create new projects, update code of existing products, test my code, do whatever Claude does as well as use the GenAI components of Grok to do research analysis for me. Only accepts messages from my number of course - it’s a linked account and accepts only from me.

keyboard_arrow_left Previous keyboard_arrow_right Next

PAI is a super power. @DanielMiessler created features on top of Claude Code that increase its efficacy by 50%... and that's a lot based on how awesome Claude Code is. Incoming FREE workshop of PAI and other tools I'm using to modify how I work in cybersecurity. Straight out of our 3.0 update of our "Red Blue Purple AI" course.

ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ @DanielMiessler

5 months ago

Super hyped to announce PAI 2.3!!! A complete rewrite of the PAI system focused around: - USER, WORK, and SYSTEM data isolation - A Continuous Learning system based on hook-based Sentiment gathering - User-based Skill Personalization ...

keyboard_arrow_left Previous keyboard_arrow_right Next

19 44 235 69K 212

🔴 (Release) The Arcanum LLM Security Context Project Today we are releasing a new AI security resource for you all! This project is a comprehensive security reference distilled from 150+ sources to help LLMs generate safer code! Enjoy! 🫶 Links Below 👇

I think we all need to be full-stack now. AI generally, but Claude Code specifically, has made it so that if you can't go from problem to a shipped solution, you're just part of a toolchain that can/will be automated. Only high-agency, full-stack people survive. And they pull away from everyone else. That’s an insane change to happen over the course of like two months.

@geoffreyhinton just captured the AI bubble debate pretty well, I think. There's not an AI tech bubble. But there is an AI investment bubble. Like I've been saying for a while now, these are very different things. And that matters for action.

This is the clearest graphic I could make on Prompt Injection. 1. Yes it's a vulnerability 2. It is the superset 3. No this does not illustrate ALL risks, just some

🧠 Did you know you can pipe into Claude Code? Like: cat logs.txt | claude -p "analyze for..." strings bin.bin | claude -p "explain X" etc...

Hackers don’t waste time — and neither should you. @Wh1t3Rh1n0 created DNS Triage, a reconnaissance tool that quickly finds information about an organization’s infrastructure, software, and even third-party services. Check out this article from Michael, where he walks you through how it works, how to install it, and how to add it to your everyday arsenal! Download it here -- github.com/Wh1t3Rh1n0/dns…

keyboard_arrow_left Previous keyboard_arrow_right Next

Everyone wants AI testing to be automated or similar to AI red teaming. Point and scan. It’s not. So much of it is a blend of web security and prompt injection. The testing is slow and manual a lot of the times. Attacks need to be hyper tailored to work for a specific businesses app. MUCH probing goes into figuring out LAYERS of evasions to bypass guardrails and classifiers. Red teaming and pivoting skills come into play using the models as vehicles. Testing is SLOWER than standard web pentesting because of the non deterministic features of the models

Catching up on my Unsupervised Learning videos by @DanielMiessler Watching his interview with @trailofbits on the architecture of their 2nd place AI Cyber Reasoning System "ButterCup 2.0"... It's SUPER validating in some areas. A few weeks back Dan and I were discussing our systems Helios and Warden and talking about how micro we have to be with the jobs for the agents. The more micro the better, the more single operation the better, otherwise context can get polluted. CONTEXT ENGINEERING IS PARAMOUNT. Check out the excellent interview here: youtube.com/watch?v=nvU0Gb…

Thread: Here are 15 FREE resources to start learning how to security test and hack generative AI 👇

Added some topics to the new course I’m building: Finding security workflows Security Prompt engineering Security Context engineering Prompt abstraction Security workflow glue - n8n, power automate, Atlassian, etc Mcp + fastapi design Benchmarking and evaluation Scaling Exposing your workflow What else?

I'm teaching an intro to cloud security workshop on July 11th. This is a pay what you can course so you can take it for free. I'll also be teaching the full version of my Breaching the Cloud course at @WWHackinFest in October. Registration links below: Workshop: antisyphontraining.com/course/worksho… Breaching the Cloud: antisyphontraining.com/course/breachi…

Pentesters: What's the coolest thing you've done with Burp Suite? 💥🖋️ New to the industry? You'll want to make sure you're comfortable with it and BB King has got you covered! THIS Friday, June 13th: antisyphontraining.com/course/worksho…

My advice to kids has changed recently and it's not to get a cybersecurity degree. Get a computer science one, with a specialization in AI or an AI solely degree. I feel like right now, the cybersecurity market is pretty rough on hiring folks in - and that's only going to get harder as AI progresses more. Having a computer science with AI focus, you can join any industry - but also be top notch in cybersecurity with new skills if you want to get into this field. I'm pretty reserved on when new things come out, and have PTSD around new cybersecurity trends that promise the world. I was the same when AI first came out - very reserved, is this going to have a huge impact? Is this really a thing. AI is much different and a monumental shift for the industry/economy and it'll only get better. The major problem we are going to run into with this is that how do we know who can really code and who's vibe coding? Pretty hard delineation/shift happening. While vibe code is cool and all, being able to actually code/develop something that's needed with augmentation from prompt coding is necessary.

You’ve received a “true positive” security alert for a Windows or Linux endpoint. This is not a drill! Your environment is under attack! This is war and you need to take rapid, decisive steps to determine: Has the endpoint been compromised? Have other systems been impacted? What actions should come next? Patterson Cake will take you through live demonstrations & hands-on labs to help you get through similar IR scenarios with confidence in our next Pay-What-You-Can Workshop: Rapid Endpoint Investigations, live THIS FRIDAY, June 6th. Register here: antisyphontraining.com/course/worksho…

Someone AI voice cloned the White House chief of staff & asked for: list of individuals the President should pardon, cash transfer, questions about President. If your team doesn’t know how to verify identity to avoid getting tricked by voice clones, now is the time to prep them.

Eric Geller @ericgeller

a year ago

Someone hacked the personal cellphone of White House chief of staff Susie Wiles and impersonated her in texts and calls (possibly with an AI voice clone) to leading Republicans and corporate executives: wsj.com/politics/polic…

9 100 237 65K 61