Uncover prompt injection, insider threats with the Tenable One Model Refusal Detection
2026-03-26T20:52:23Z•a8ee016772a042aa2638ddd1c2710fd1e1b2b2f16c6ce523e2c963785c9afcca
AI governanceAI securityAppSecCNAPPCVE-2025-61757CVE-2026-21514CVE-2026-21992LLM securityMicrosoft WordOLE bypassOracle Identity ManagerOracle Web Services ManagerTenable Hexa AITenable Oneacceptable use policyagentic AIexposure managementinsider threatmodel refusal detectionprompt injectionremote code executionsupply chain riskvulnerability management
What happened
This Tenable blog feed centers on AI-focused security and exposure management and highlights multiple high-impact vulnerabilities. Tenable announced Model Refusal Detection in Tenable One AI Exposure to turn LLM refusals into early-warning signals for prompt injection, insider threats, and adversarial prompting. The posts also introduce Tenable Hexa AI (an agentic engine for orchestrating exposure remediation) and recommend AI governance, AUPs, and unified exposure management to address new language-based attack surfaces. Notable vulnerability coverage includes CVE-2026-21992 (Oracle Identity
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- tenable_blog
- Record identifier
- a8ee016772a042aa2638ddd1c2710fd1e1b2b2f16c6ce523e2c963785c9afcca
- Enrichment time
- 2026-03-26T20:52:23Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.