Executive Summary
CNCERT warned that OpenClaw AI agents have vulnerabilities enabling indirect prompt injection via untrusted content, leading to data leaks or unauthorized actions.
What Happened
CTI Labs highlighted a CNCERT critical warning that OpenClaw AI agents contain vulnerabilities enabling indirect prompt injection via untrusted content, which can cause agents to follow attacker-specified instructions instead of intended workflows.[11] The advisory warns that compromised agents may leak sensitive organizational data or trigger unauthorized actions when processing poisoned web pages, documents, or emails within business environments.[11]
Affected Products
- OpenClaw
Vulnerability / Attack Technique
Indirect prompt injection
Technical Mechanism
Poisoned web pages, documents or emails make agents follow attacker instructions.
AI Security Category
LLM01:2025 Prompt Injection
Part of the Prompt Injection topic.
Relevant Frameworks
RealGround Analysis
A national CERT warning on a popular agent framework. RealGround analysis: organizations using OpenClaw should restrict its data and tool access.
Recommended Controls
- Update OpenClaw
- Limit OpenClaw tool and data access
Related RealGround Service
AI Agent Business Logic Audit · or start with the free AI Agent Security Scanner.
Sources
CTI Labs (via CNCERT advisory summary): https://www.linkedin.com/posts/cti-labs-io_aisecurity-promptinjection-dataleak-activity-7439234378960060416-LpuF
Summary and analysis written by RealGround from the source above; last updated 2026-10-18.