OpenClaw integrates Tencent AI tool into security checks
OpenClaw has added Tencent's AI-Infra-Guard tool to its security checks for the ClawHub registry. The new combined review process enhances detection of malicious inputs.
OpenClaw, the maintainer of the ClawScan review tool and ClawHub registry, has integrated Tencent's AI-Infra-Guard (AIG) artificial intelligence tool into its security checks. This integration enhances the security screening for all AI skills and plugins uploaded to ClawHub.
The new system combines the separate analyses from ClawScan's AIG and Nvidia's SkillSpector tools. An AI judge then evaluates the findings from both tools alongside the uploaded files, thereby improving overall security.
OpenClaw and Tencent tested the method using the SkillTrustBench dataset, a benchmark developed by Tencent and the Chinese University of Hong Kong, Shenzhen. Tests indicated that the combined process matched 86.9% of the benchmark labels and correctly classified 98.6% of malicious cases.
This update occurs amidst broader scrutiny of AI agent extensions. The integration aims to significantly improve the detection of malicious AI applications and plugins.