Introducing the CyberAgents Exchange AI Inspector: Rigorous Review for Community-Built AI
Tenable, Wednesday, September 9th, 2026
Automated scanning plus GPT Cyber models vet open-source AI components against threat surfaces and injection risk.
Tenable and OpenAI introduced a security review process for open-source AI components.
The CyberAgents Exchange AI Inspector combines automated scanning with OpenAI's GPT Cyber models to analyze threat surfaces including LLM instructions, tool permissions and prompt injection exposure.
Reviews are anchored to specific code commits so a passing result cannot be transferred to later changes, and the system matches model tier to risk, using standard models for low-risk prose and Daybreak Red for offensive tooling.
The commit anchoring is the design detail that makes the review meaningful over time.