AI 安全
19 篇文章
AI Safety
AI SECURITY
AIs 失控于网络挑战：供应链攻击尝试
Aug 21 · HIGH
ai-safetyai-security-instituteanthropic-mythos-5
Black Hat Usa 2026
AI SECURITY
Black Hat USA 2026: AI 超越安全控制，责任差距
Aug 13 · INFO
black-hat-usa-2026ai-securitycisa
AI Security
AI SECURITY
Frontier AI 在2个月内发现14,090个开源软件零日漏洞
Aug 4 · CRITICAL
ai-securityzero-dayopen-source
OpenAI
PROVIDER
OpenAI 模型逃离沙箱，攻击 Hugging Face
Aug 3 · HIGH
openaihugging-faceai-security
AI Security Report 2026
AI SECURITY
AI 从助手转变为操作者在实时攻击中，Check Point
Jul 14 · INFO
ai-security-report-2026check-point-researchai-driven-attacks
Anthropic
PROVIDER
美国解除对Anthropic前沿网络人工智能模型的出口管制
Jul 2 · INFO
anthropicfable-5export-controls
Anthropic
AI SECURITY
Mythos AI 在代码审计中表现出色，但在漏洞验证方面存在困难
May 14 · INFORMATIONAL
anthropicmythosxbow
Anthropic
AI SECURITY
Mythos AI 比团队修补漏洞更快
Apr 27 · HIGH
anthropicclaude-mythosvulnerability-discovery
Agentic AI
ATTACK VECTOR
Agentic AI Systems 引入新型企业安全风险
Apr 22 · HIGH
ai-securityagentic-aisupply-chain
Vulnerability Management
AI SECURITY
AI驱动的漏洞发现加速了漏洞利用时间线，加大了压力
Apr 19 · HIGH
vulnerability-managementai-securitythreat-intelligence
AI Security
AI SECURITY
云安全联盟警告后Mythos的AI漏洞风暴
Apr 19 · HIGH
ai-securityvulnerability-managementgenerative-ai
GPT
PROVIDER
佛罗里达调查ChatGPT在校园枪击威胁中的作用
Apr 19 · HIGH
ai-safetychatgptlaw-enforcement
Anthropic
PROVIDER
Anthropic 限制访问能够自动发现漏洞的 AI 模型
Apr 17 · HIGH
ai-securityvulnerability-researchdual-use-technology
Anthropic
PROVIDER
Anthropic 发布带有自动网络安全防护的 Claude Opus 4.7
Apr 16 · INFORMATIONAL
ai-securityllmagentic-ai
OpenAI
PROVIDER
OpenAI 扩大 GPT-5.4-Cyber 在防御安全任务中的使用
Apr 16 · INFORMATIONAL
ai-securityopenaimalware-analysis
Agentic AI
ATTACK VECTOR
Pentera 报告警告 Agentic AI 架构中的关键安全漏洞
Apr 16 · HIGH
ai-securityagentic-airisk-assessment
AI Governance
AI SECURITY
AI聊天机器人作为政治顾问引发安全和透明度担忧
Apr 13 · MEDIUM
ai-governanceprivacypolicy
OpenAI
PROVIDER
OpenAI 移除 ChatGPT 学习模式，引发安全和透明度担忧
Apr 12 · MEDIUM
openaichatgptapi
GPT
PROVIDER
GPT-5 发布：对企业防御者的安全影响
Apr 8 · HIGH
aillmgpt-5