Dreadnode’s Offensive AI Capabilities on Display in Vegas During Black Hat, Def Con, and the AI Security Forum

August 5, 2025—Las Vegas, NV—Dreadnode is heading to Las Vegas August 4-10 for Hacker Summer Camp. Alongside the team, Co-Founders Will Pearce and Nick Landers will be onsite all week dropping new research, showcasing the latest offensive AI platform updates, and bringing “shiny rocks” that demonstrate the capabilities of models executing offensive security tasks.

Reach out at dreadnode.io/contact-us or on socials (LinkedIn, X) to request a meeting.

Advance Offensive Security with the Dreadnode Platform

Strikes is Dreadnode’s offensive AI platform for deploying, evaluating, and observing agents. Dreadnode provides the only platform that enables security teams to automate offensive security tasks, assess AI systems, and scale capabilities through advanced agents and red teaming tools.

Schedule a demo to see the latest platform updates, including:

• Scale offensive security testing using our new agents that find and exploit vulnerabilities, run network operations end-to-end, and test web applications. Reduce time and manual effort on repetitive tasks.
• Execute AI red team assessments against single (and multi!) modal models using the Dreadnode SDK’s AIRT tools. Evaluate and report results with the Strikes UI.
• Build and optimize observable agents that you can evaluate for performance. Develop agents using the Dreadnode SDK with full observability into agent behavior.
• Track and compare agent performance via visual analytics with robust charting, graphing, and comparison tools that eliminate the need for custom reporting.
• Analyze multi-modal model outputs including image, video, and audio analysis during agent workflows. Test beyond text based assessments.
• Control your data and deploy a fully featured, private instance of Strikes for your team, in your environment.

Research and Shiny Rocks

The team is bringing a stacked lineup of research and novel findings to showcase, centered around the evaluation and deployment of agents. Meet with us to discuss our findings across reversing, multi-step network attacks, bug bounty, threat intel, and more.

Additionally, led by Dreadnode Principal Research Engineer Shane Dawson, we will be publishing a research paper that compares LLMs and humans at judging the success of a penetration testing agent. It is a North Star for autonomous processes and, notably, the judge outputs of these rubrics can be used as a reward-signal in order to train agents that consider multiple objectives necessary for production agents.

Black Hat AI Summit—Securing Deterministic vs. Non-Deterministic AI Systems: A New Paradigm for Cybersecurity

Tuesday, August 5 | 2:25-3:25 pm PT | Oceanside A, Level 2

Will is moderating a panel during Black Hat’s AI Summit on the urgent need for a paradigm shift around securing non-deterministic AI systems versus traditional deterministic systems. Hear insights on how to safeguard the next generation of intelligent systems from Tori Westerhoff,  Director in AI Safety & Security Red Teaming at Microsoft; Jonathan Dambrot, CEO at Cranium AI; Monzy Merza, Co-Founder and CEO at Crogl; and Clay Lin, Senior Advisor at World Bank Group. Register here.

Las Vegas AI Security Forum—”Welcome to the Offensive AI Party” Session and Cyber Agent Building Workshop

Thursday, August 7 | Palms Casino Resort, Las Vegas

Will and Nick will be participating in the Las Vegas AI Security Forum, giving a talk on offensive AI and leading an agentic cyber workshop. We are also facilitating a women in AI security meetup alongside the organizers. Apply to attend here, or reach out to tori@dreadnode.io.

Talk: Welcome to the Offensive AI Party | Will Pearce | 2-2:30 pm PT

Initially called out in the Biden Executive Order, the capabilities of models to execute offensive security tasks has been hotly debated. It is now 2025, and several examples of models not just performing these tasks, but excelling at them have been shown. This talk will discuss Dreadnode’s findings across the development and experimentation with offensive agents. This technology is multi-use at its core, and offers uplift to all sides.

Workshop: Building Cyber Agents | Nick Landers | 3-4 pm PT

Everyone wants to build cyber agents, but they don’t have the building blocks to go from PoC to production. In this workshop attendees will learn how to use our open-source frameworks to deploy agents on real world offensive tasks.

Def Con Bug Bounty Village—Misaligned: AI Jailbreaking Panel

Dreadnode Staff AI Security Researchers Ads Dawson is participating in a panel hosted by Jason Haddix and the BASI group, the notorious AI model jailbreak group led by Pliny the Prompter. The panel will discuss jailbreaking war stories, techniques, and opportunities to get into AI hacking for profit. More information on the Bug Bounty Village here.

You can also find Ads at various bug bounty and CTF competitions during the week. He will be leveraging his bug bounty harness to showcase the capabilities of hacking with AI at scale.