Cyber Dialogues by Palo Alto Networks

Agentic AI: The Autonomous Threat and the New Frontier of Defense

Palo Alto Networks Season 1 Episode 11

Guest: Andrew Milroy, Distinguished Analyst and Founder of Veqtor8

In this episode of Cyber Dialogues, we explore the rapid emergence of Agentic AI, autonomous systems that can perceive, reason, plan, and execute actions independently. 

Host Sharmin Jassal speaks with renowned analyst Andrew Milroy about the unprecedented security challenges and defensive opportunities this technology presents.

Andrew explains that the current rush to adopt AI for productivity is creating an exponentially larger attack surface, a pattern he likens to the early days of Cloud and Mobile adoption where security was an afterthought. He details two critical and unique vulnerabilities:

  1. Model Poisoning and Prompt Injection: How malicious data or disguised commands can manipulate an autonomous agent's behavior, leading to unauthorized actions or sabotage.
  2. Detection Difficulty: Why the adaptive, constantly changing nature of Agentic AI makes it difficult for static security rules to detect anomalous or compromised activity.

The conversation then pivots to the offensive power of these agents, discussing how they can supercharge threats with speed and scale, making phishing attacks almost indistinguishable from legitimate communications. Finally, Andrew offers an optimistic outlook, highlighting the opportunity for enterprises to adopt Agentic SecOps - autonomous defense systems capable of instantly detecting, responding to, and shifting security posture against sophisticated, evolving threats.

Key Takeaways:

  • The AI Vulnerability: AI adoption is currently being driven by enthusiasm for efficiency, often ignoring security risks, leading to a massive expansion of the attack surface.
  • The Attacker's Edge: Agentic AI enables attackers to create fully automated, Cunning, and adaptive threats that can respond dynamically to human attempts at validation.
  • The Defender's Opportunity: Security Operations Centers can leverage Agentic AI to automate threat hunting, reduce false positives, and achieve response speeds that were previously impossible, effectively turning defense into an autonomous function.
  • Market Transformation: Andrew believes Agentic AI will disrupt the entire IT industry, potentially marking a shift away from standardized SaaS platforms toward highly customized, data-proximate agent architectures.


  • Connect with Andrew Milroy on LinkedIn: 

https://www.linkedin.com/in/andrewmilroy/