½ÃÀ庸°í¼­
»óǰÄÚµå
1748021

»çÀ̹ö º¸¾ÈÀÇ ¿¡ÀÌÀüÆ® AI : »çÀ̹ö º¸¾È ¼³°èÀÚ¸¦ À§ÇÑ ÀÔ¹® °¡À̵å

Agentic AI in Cybersecurity: A Primer Guide for Cybersecurity Architects

¹ßÇàÀÏ: | ¸®¼­Ä¡»ç: IDC | ÆäÀÌÁö Á¤º¸: ¿µ¹® 17 Pages | ¹è¼Û¾È³» : Áï½Ã¹è¼Û

    
    
    



¡Ø º» »óǰÀº ¿µ¹® ÀÚ·á·Î Çѱ۰ú ¿µ¹® ¸ñÂ÷¿¡ ºÒÀÏÄ¡ÇÏ´Â ³»¿ëÀÌ ÀÖÀ» °æ¿ì ¿µ¹®À» ¿ì¼±ÇÕ´Ï´Ù. Á¤È®ÇÑ °ËÅ並 À§ÇØ ¿µ¹® ¸ñÂ÷¸¦ Âü°íÇØÁֽñ⠹ٶø´Ï´Ù.

»çÀ̹ö º¸¾ÈÀÇ ¿¡ÀÌÀüÆ® AI¿¡ ´ëÇÑ IDC Perspective¿¡¼­´Â ½Ã½ºÅÛÀÌ ÀνÄ, Ãß·Ð, ÇൿÀ» ÅëÇØ ÀÚÀ²¼ºÀ» ¹ßÈÖÇÒ ¼ö ÀÖµµ·Ï ÇÏ´Â ¿¡ÀÌÀüÆ® AIÀÇ ¿ªÇÒÀ» °­Á¶Çϸ鼭 »çÀ̹ö º¸¾È¿¡¼­ AIÀÇ ÁøÈ­¿¡ ´ëÇØ »ìÆìº¾´Ï´Ù. ÀÌ ¹®¼­¿¡¼­´Â ¼öµ¿ ¿î¿µ¿¡¼­ µ¿Àû ÀÚÀ²¼ºÀ¸·ÎÀÇ ¹ßÀü °úÁ¤À» ¼³¸íÇϸç ÀÚÀ²¼ºÀ» ´Þ¼ºÇÏ´Â µ¥ ÀÖ¾î ´Ü¼ø¼º, º¸¾È Ç÷§Æû, Ç¥ÁØ IT ¾ÆÅ°ÅØÃ³ÀÇ Á߿伺À» °­Á¶ÇÕ´Ï´Ù. ¶ÇÇÑ »çÀ̹ö º¸¾È¿¡¼­ ¾ÈÀüÇϰí À±¸®ÀûÀÎ AI ¹èÆ÷¸¦ º¸ÀåÇϱâ À§ÇØ ¼³¸í °¡´É¼º ¹× °øÁ¤¼º°ú °°Àº ½Å·ÚÇÒ ¼ö ÀÖ´Â AI ¿ä¼ÒÀÇ ÅëÇÕ¿¡ ´ëÇØ¼­µµ ³íÀÇÇÕ´Ï´Ù. "¿¡ÀÌÀüÆ® AI Ç÷§ÆûÀº Ãß·Ð ¿£ÁøÀÌ Á¡Á¡ ´õ Á¤±³ÇØÁö°í È®·üÀû ÀÇ»ç °áÁ¤À» À§ÇÑ º¸»ó ¼ö´ÜÀÌ °³¼±µÊ¿¡ µû¶ó º¸¾È ȯ°æÀÇ º¯È­¿¡ µû¶ó ¹Ýº¹Çϰí Á¶Á¤Çϱ⠽ÃÀÛÇÕ´Ï´Ù. AI ¿¡ÀÌÀüÆ®´Â ŽÁö ¹× Á¶»çºÎÅÍ ´ëÀÀ ¹× ¼öÁ¤¿¡ À̸£±â±îÁö ´ëºÎºÐÀÇ ´Ü¼øÇÏ°í ¹Ýº¹ÀûÀÎ SOC ÇÁ·Î¼¼½º¸¦ »ç¶÷ÀÇ °³ÀÔÀ» ÃÖ¼ÒÈ­Çϸ鼭 ó¸®ÇϹǷΠMTTI¿Í MTTRÀÌ ´õ¿í ´ÜÃàµË´Ï´Ù. Çǵå¹é ·çÇÁ¸¦ ÅëÇØ AI ¿¡ÀÌÀüÆ®´Â ÀÌÀü¿¡ ¹ß°ßµÈ À§ÇùÀ» ±â¹ÝÀ¸·Î »õ·Î¿î ŽÁö ¹× ´ëÀÀÀ» »ý¼ºÇÒ ¼ö ÀÖ½À´Ï´Ù. Àΰ£Àº »õ·Î¿î À§Çù¿¡ ´ëÇÑ Ç÷§ÆûÀÇ ÀûÀÀÀ» °¨µ¶Çϰí IT ¾ÆÅ°ÅØÃ³ÀÇ º¯È­¿¡ ÀûÀÀÇϸ鼭 º¹¿ø·ÂÀ» À¯ÁöÇÏ´Â Àü·«Àû ¿ªÇÒÀ» ¼öÇàÇÕ´Ï´Ù. À̸¦ Çö¸íÇÏ°Ô »ý°¢ÇÏ´Â ¹æ¹ýÀº "½Ã½ºÅÛ¿¡ Àΰ£ÀÇ ÆÇ´Ü°ú °³ÀÔÀ» ³Ö´Â °Í"ÀÌ ¾Æ´Ï¶ó "»ç¶÷ÀÌ ½Ã½ºÅÛÀ» ¸ð´ÏÅ͸µÇÏ´Â °Í"ÀÔ´Ï´Ù." - Frank Dickson, IDC º¸¾È ¹× ½Å·Ú ºÎ¹® ±×·ì ºÎ»çÀå

ÁÖ¿ä ¿ä¾à

»óȲ °³¿ä

  • »çÀ̹ö º¸¾ÈÀÇ ¿¡ÀÌÀüÆ® AI·ÎÀÇ ÁøÈ­

±â¼ú ±¸¸ÅÀÚ¸¦ À§ÇÑ Á¶¾ð

  • ¿¡ÀÌÀüÆ® AI¿Í ÄÚÆÄÀÏ·µÇü AIÀÇ »ç¿ë »ç·ÊÀÇ Â÷ÀÌÁ¡ ÀÌÇØÇϱâ
  • ¿¡ÀÌÀüÆ® AI Á¢±Ù ¹æ½Ä °èȹ
  • ¿¡ÀÌÀüÆ® AIÀÇ ¹ßÀü
  • ½Å·ÚÇÒ ¼ö ÀÖ´Â AI ¿ä¼Ò
    • °³¿ä °¡´É¼º - ¿¡ÀÌÀüÆ® AI ½Ã½ºÅÛÀÌ Àΰ£ÀÌ ÀÌÇØÇÒ ¼ö ÀÖ´Â ÇüÅ·ΠÀÇ»ç°áÁ¤ÀÇ ÀÌÀ¯¸¦ ¸íÈ®ÇÏ°Ô ¼³¸íÇÒ ¼ö ÀÖ´Â ´É·Â
      • ±×°ÍÀÌ Áß¿äÇÑ ÀÌÀ¯
      • Áú¹®ÇØ¾ß ÇÒ °Íµé
    • °øÁ¤¼º - ¿¡ÀÌÀüÆ® AI ½Ã½ºÅÛÀÌ ¼­·Î ´Ù¸¥ Àα¸Åë°èÇÐÀû ¶Ç´Â »çȸÀû ±×·ì °£¿¡ ÆíÇâµÇ°Å³ª Â÷º°ÀûÀÎ °á°ú¸¦ »ý¼ºÇÏÁö ¾Êµµ·Ï º¸ÀåÇÕ´Ï´Ù.
      • ±×°ÍÀÌ Áß¿äÇÑ ÀÌÀ¯
      • Áú¹®ÇØ¾ß ÇÒ °Íµé
    • Åõ¸í¼º - µ¥ÀÌÅÍ ¼Ò½º, ¸ðµ¨ ¾ÆÅ°ÅØÃ³, ÀÇ»ç°áÁ¤ ÇÁ·Î¼¼½º µî ¿¡ÀÌÀüÆ® AI ½Ã½ºÅÛÀÇ ±¸Ãà, ÈÆ·Ã, ¹èÆ÷ ¹æ¹ý¿¡ ´ëÇÑ °³¹æ¼º
      • ±×°ÍÀÌ Áß¿äÇÑ ÀÌÀ¯
      • Áú¹®ÇØ¾ß ÇÒ °Íµé
    • Á¤È®Çϰí ÀûÀýÇÑ - ¿¡ÀÌÀüÆ® AI ½Ã½ºÅÛÀÌ ±â´ÉÀû ¸ñÀû¿¡ µû¶ó ÀÛµ¿ÇÏ°í ¼³°èµÈ ¹üÀ§ ³»¿¡¼­ »ç¿ëµÇµµ·Ï º¸ÀåÇÕ´Ï´Ù.
      • ±×°ÍÀÌ Áß¿äÇÑ ÀÌÀ¯
      • Áú¹®ÇØ¾ß ÇÒ °Íµé
    • ±â¿ø°ú °èº¸ - AI ¶óÀÌÇÁ»çÀÌŬ Àü¹Ý¿¡ °ÉÃÄ µ¥ÀÌÅÍ, ¸ðµ¨, ÀÇ»ç°áÁ¤ÀÇ ±â¿ø°ú ÁøÈ­ ÃßÀû
      • ±×°ÍÀÌ Áß¿äÇÑ ÀÌÀ¯
      • Áú¹®ÇØ¾ß ÇÒ °Íµé
    • Àû´ëÀû °ß°í¼º - ¿¡ÀÌÀüÆ® AI ½Ã½ºÅÛÀÌ ¼ÓÀ̱â À§ÇØ ¼³°èµÈ Á¶ÀÛÀ̳ª °ø°Ý¿¡ ÀúÇ×ÇÒ ¼ö ÀÖ´Â ´É·Â
      • ±×°ÍÀÌ Áß¿äÇÑ ÀÌÀ¯
      • Áú¹®ÇØ¾ß ÇÒ °Íµé
  • À̸¦ ¿¡ÀÌÀüÆ® AI Àü·«¿¡ ÅëÇÕÇÏ´Â °Í

Âü°íÀÚ·á

  • °ü·Ã Á¶»ç
  • ¿ä¾à
ksm

This IDC Perspective on agentic AI in cybersecurity explores the evolution of AI in cybersecurity, emphasizing the role of agentic AI, which enables systems to exhibit agency through perception, reasoning, and action. The document outlines the progression from manual operations to dynamic autonomy, highlighting the importance of simplicity, security platforms, and standard IT architecture in achieving autonomy. It also discusses the integration of trustworthy AI elements, such as explainability and fairness, to ensure safe and ethical AI deployment in cybersecurity."Agentic AI platforms begin to iterate and adjust to changes in the security environment as reasoning engines become increasingly sophisticated and the compensating measures for probabilistic decision-making improve. AI agents handle most menial and repetitive SOC processes, from detection and investigation to response and remediation, with minimal human intervention, further reducing MTTI and MTTR. A feedback loop enables AI agents to create new detections and responses based on previously seen threats. Humans play a strategic role, overseeing the platform's adaptation to emerging threats and adjusting to changes in the IT architecture while maintaining resilience. A smart way to think of this is putting the 'human on the loop' as opposed to putting the 'human in the loop'." - Frank Dickson, group vice president, Security and Trust, IDC

Executive Snapshot

Situation Overview

  • The Progression to Agentic AI in Cybersecurity

Advice for the Technology Buyer

  • Knowing the Difference Between Agentic AI Versus Copilot Use Cases
  • Planning for Your Agentic AI Approach
  • Agentic AI Progression
  • Elements of Trustworthy AI
    • Explainability - The Ability of an Agentic AI System to Articulate the Reasoning Behind Its Decisions in a Way That Humans Can Understand
      • Why It Matters
      • Questions to Ask
    • Fairness - Ensuring That Agentic AI Systems Do Not Produce Biased or Discriminatory Outcomes Across Different Demographic or Social Groups
      • Why It Matters
      • Questions to Ask
    • Transparency - Openness About How the Agentic AI System Is Built, Trained, and Deployed, Including Data Sources, Model Architecture, and Decision-Making Processes
      • Why It Matters
      • Questions to Ask
    • Accurate and Appropriate - Ensuring the Agentic AI System Performs to the Functional Purpose and Is Used Within the Scope It Was Designed for
      • Why It Matters
      • Questions to Ask
    • Provenance and Lineage - Tracking the Origin and Evolution of Data, Models, and Decisions Throughout the AI Life Cycle
      • Why It Matters
      • Questions to Ask
    • Adversarial Robustness - The Ability of an Agentic AI System to Resist Manipulation or Attacks Designed to Fool It
      • Why It Matters
      • Questions to Ask
  • Integrating These into an Agentic AI Strategy

Learn More

  • Related Research
  • Synopsis
»ùÇà ¿äû ¸ñ·Ï
0 °ÇÀÇ »óǰÀ» ¼±Åà Áß
¸ñ·Ï º¸±â
Àüü»èÁ¦