Expired50% confidenceFactExact time
Security issues in large language models currently lack a standardized classification system analogous to CVE, and there is no unified standard for assessing vulnerability severity.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/12/2026
First Seen
Valid until: 10/10/2026(expired)
Sources
Claude Mythos预览版漏洞激增：大模型安全性如何保障？
hackernewshackernews7/3/2026
Related Claims
UnverifiedThe AI safety field still lacks a standardized vulnerability disclosure framework similar to CVE.79% similarUnverifiedVulnerability reports lacking reproducible evidence (including a PoC, a clearly defined range of affected versions, and independently verifiable technical details) should not be assigned an official CVE number.71% similarUnverifiedNo unified LLM safety evaluation standard that is as widely cited as MMLU has yet emerged.69% similarUnverifiedThere is currently no universally recognized standard for evaluating prompt quality in the industry.68% similarUnverifiedRepresentative works in safety evaluation datasets include AdvBench, HarmBench, and SALAD-Bench, and a unified standard like MMLU has yet to be established65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/490540API
curl https://kongchang.com/api/v1/knowledge/claims/490540MCP
get_claim(id=490540)