What happened
An OWASP Cornucopia companion edition 'LLMX / Large Language Models' (v1.0) is live at cornucopia.owasp.org, presenting LLM threat-modeling cards mapped to OWASP LLM Top 10 (LLM01:2026), MITRE ATLAS, OWASP AISVS and AITG, STRIDE, PHANTOM-B and CWE. The page is confirmed real and OWASP-hosted, but no publication date could be recovered from the page or search results (no [extracted-metadata] date header), so the date is uncertain and may predate the reporting window.
Why it matters
An OWASP threat-modeling companion for large language models that aligns LLM threats with the 2026 LLM Top 10, ATLAS, and the AI Security Verification Standard / AI Test Guide — useful for structured LLM threat modelling. Date unconfirmed, so included for QA adjudication.
Action needed
Practitioners doing LLM threat modelling may use the LLMX cards to map LLM Top 10 and ATLAS risks; QA should confirm whether this edition was released in the reporting window.