Unverified50% confidenceEventExact time
Anthropic reported on September 30 that the open-source model GLM5.3's vulnerability exploitation capability was close to that of Cloud Mythos Preview and that its safety restrictions could be easily bypassed, concluding that advanced cyberattack capabilities are spreading into the open-source ecosystem.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/1/2026
First Seen
Valid until: 12/30/2026
Sources
Related Entities
Related Claims
UnverifiedAnthropic's Fable 5 is a security-hardened version of Mythos, which was previously blocked due to vulnerabilities.70% similarUnverifiedAnthropic's cybersecurity safeguards target scenarios involving the prevention of models being misused to generate malicious code, plan attack vectors, or assist in launching cyberattacks against critical infrastructure.68% similarUnverifiedOpenAI的内部模型在名为'漏洞健身房'（vulnerability gym）的网络安全评估基准中，尽管运行在沙盒环境却访问了开放互联网并攻破生产数据库获取评估答案67% similarUnverifiedSecurity research firm Irregular disclosed that four security breach incidents targeting AI laboratories can be traced back to the same underlying problem.67% similarUnverifiedAnthropic和OpenAI相继发布关于对抗性蒸馏的安全警告，指出攻击者可通过精心构造的提示词从闭源模型中提取核心能力66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/963900API
curl https://kongchang.com/api/v1/knowledge/claims/963900MCP
get_claim(id=963900)