Researchers say an AI-powered code scanner traced untrusted data across layers of OpenClaw, exposing exploitable weaknesses including SSRF, authentication bypass, and path traversal.

CSO Online offers insights into cybersecurity, risk management, and IT leadership, providing articles, reports, and expert analysis to help security professionals navigate the evolving threat landscape and protect their organizations from cyber attacks. By exploring CSO Online's curated content, CISOs, security managers, and IT executives can learn about threat intelligence, security frameworks, and incident response strategies for building resilient cybersecurity programs. Whether you're defending against ransomware, phishing attacks, or insider threats, CSO Online offers resources to strengthen your security posture and safeguard your digital assets.

CSO Online

Endor Labs discovered six high-to-critical security vulnerabilities in OpenClaw, an open-source AI agent framework. Using an AI-powered static analysis engine that traces untrusted data flows across software layers, researchers identified SSRF bugs in a gateway component, Urbit Authentication, and an Image Tool; a missing webhook verification flaw in a Telnyx handler; an authentication bypass in Twilio webhook functionality; and a path traversal issue in browser upload handling. Working proof-of-concept exploits were published for each flaw. OpenClaw has since released patches and security advisories. The research highlights how AI-driven SAST can maintain context across complex data transformations to surface vulnerabilities that traditional static analysis tools may miss.

Six flaws found hiding in OpenClaw’s plumbing