Overview

RunSybil builds AI-native security tooling that continuously tests live systems. Its flagship product, the Sybil AI agent, performs autonomous black-box penetration testing against applications and infrastructure by interacting with real interfaces rather than relying on static code analysis. Sybil reasons like a hacker, chaining small weaknesses to uncover privilege escalations or cross-tenant data access, and it automatically documents findings for fast remediation. Unlike traditional scanners, it operates without human input and provides security feedback on every pull request, helping teams reduce false positives and move away from infrequent bug bounty programs. The company targets a subscription-based model for startups and large enterprises, including names like Cursor, Notion, and Turbopuffer, and raised funding to scale R&D and go-to-market efforts. Its goal is to automate hacker-like intuition to continuously improve an organization’s security posture.