PPhying中文
Curated security research, vulnerabilities, advisories and tools for practitioners.

AI Pentesting Without the Noise: Hackbots and the Verifier

ethiack.com | research | #ai-security | #ai-pentesting | #verifier | #hackbot | #hallucination | #ethiack | #owasp-benchmark
📄 Raw .md⬇️ Download .md🔗 Original ↗

Summary

Ethiack explains how its hackbot pairs autonomous vulnerability discovery with a Verifier to curb AI hallucinations, and how OWASP Benchmark testing validates precision in black-box conditions.
Published
Collected

Related coverage

blog · ethiack.com
Ethiack Raises a €4 Million Funding Round to Develop AI-Powered Hackbots
Ethiack raises €4M seed to scale R&D on its AI-powered Hackbot for continuous pentesting; it reports $1M ARR, 50+ clients in 9 countries, and plans for UK and US expansion.
blog · hackerone.com
What We’ve Learned from 5 Months of Hackbot Activity
Five months after AI hackbots joined HackerOne, XBOW topped the US leaderboard and Burp Suite added MCP support. Hallucinated vulnerability reports remain the key challenge.
research · ethiack.com
Evaluating Pentesting Agents for the Real-World Part 2
Ethiack's EthiBench part 2 evaluates AI pentesting agents beyond single runs, covering significance under budget limits, cumulative detection, temporal behavior, and cost-effective target subsets.
research · ethiack.com
Evaluating Pentesting Agents for the Real-World
Ethiack launches EthiBench, an open evaluation protocol for AI pentesting agents using realistic targets, structured ground truth with LLM-as-a-judge matching, and richer metrics than success rates.
research · ethiack.com
Hackian and the Ghosts in Your Business Logic
Ethiack's Hackian agent exploited business-logic flaws scanners miss, chaining a SonarQube branch-metadata permission gap and a second access-control issue to reach private projects it shouldn't see.
research · ethiack.com
One-click RCE on OpenClaw in under 2 hours with an Autonomous Hacking Agent
Ethiack's Hackian agent found a one-click account-takeover-to-RCE chain in OpenClaw fully autonomously in under two hours, leaking the Gateway Control UI's auth token from an attacker-controlled site.
Back