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
Research Assistance
Learn how to effectively navigate the market research process to help guide your organization on the journey to success.
Download eBook