New security tools emerge to vet AI infrastructure and agents

New security tools are being developed to address the risks associated with the rapid adoption of AI agents and infrastructure. Tencent’s Zhuque Lab has released AI-Infra-Guard, an open-source security scanner designed to fingerprint running services like Ollama and vLLM. The tool checks services against over 1,600 known vulnerabilities (CVEs) and evaluates agent skills across 14 risk categories using the SkillTrustBench dataset. The platform is currently utilized by organizations including ICBC, Lenovo, and vivo.

In a separate development, Tenable is collaborating with OpenAI to launch the CyberAgents Exchange AI Inspector. This security review process aims to vet AI agents, skills, and multi-agent playbooks within Tenable’s CyberAgents Exchange. The inspector combines OpenAI’s GPT cyber models with Tenable’s security expertise to help enterprise security teams safely deploy agentic AI components. The capability is expected to be available in September 2026.