SecureAITransformation-WhatWeCanDoNowandintheFuture

3.0 2025-05-11 61 0 4907 KB 14 页 PDF
侵权投诉
SecureAITransformation-WhatWeCanDoNowandintheFuture
SecureAITransformation-WhatWeCanDoNowandintheFuture
SecureAITransformation-WhatWeCanDoNowandintheFuture
SecureAITransformation-WhatWeCanDoNowandintheFuture
SecureAITransformation-WhatWeCanDoNowandintheFuture
摘要:

PRESENTED BYSecure AI transformation: What we can do now and in the futureEfim Hudis, VP Microsoft Security Neta Haiby, Director of AI Security, Microsoft Security2AI is shifting business Every Every Every 3GenAI apps are a blackbox, versatile, probabilistic, and not deterministicHave high connectivity and autonomyUse natural language and can be manipulated4You engage with AI in various ways YOU USE ITYOU BUILD ITLow code/No Code/Code5Generative AI introduces new attack surfaces and risksPrompt injectionJailbreakData leakage in promptsRisky contentAI regulationsTraining dataPromptsResponsesPrompt integrationPlug-insOrchestrationModelsApplicationIdentityEndpointsDataCloud6Generative AI threat mapMITRE ATLASOWASP Top 10 for LLMMSRC AI Bug BarOWASP Top 10 for MLAI Usage SecurityUser interaction with generative AI-based appsSensitive information disclosureShadow IT/harmful third-party LLM-based app or pluginAI ApplicationSecurityGenerative AI-based app lifecycle UPIA/XPIA Intent-breakingDa

展开>> 收起<<
SecureAITransformation-WhatWeCanDoNowandintheFuture

共 14 页,预览3页

还剩11页未读, 继续阅读

声明:企商查报告文库所有资源均是客户上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作商用。
作者: 分类: 属性:14 页 大小:4907 KB 格式:PDF 时间:2025-05-11

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 3
客服
关注