Agentic Boom Exposes Gap in AI Security and Governance

Daryan “D” Dehghanpisheh, director of AI security, Palo Alto Networks

Agentic artificial intelligence adoption gained traction quicker than generative AI models upon release, signaling that organizations don’t just want content generation tools but agents that take action.

Daryan “D” Dehghanpisheh, director of AI security at Palo Alto Networks, said agent adoption is accelerating because agents directly drive productivity and revenue, making them unavoidable across enterprise environments. That speed, however, is outpacing governance and security readiness.

But the lack of an AI security governance framework is still a challenge. “Every vendor who comes in with an AI tool is actually being questioned about how performant it is, how secure it is, how safe it is, how governed it is. Security, safety and governance are three distinctly different things. They’re related, but they are mutually exclusive,” he said.

In this video interview with Information Security Media Group, Dehghanpisheh also discussed:

Securing agent actions through least-privilege controls;
Model security, testing and insider risk;
The rise of universal AI vulnerabilities and coordinated disclosure gaps.

Dehghanpisheh is a senior executive with more than 20 years of experience in technology and strategy, leading teams in industries ranging from telecom to finance. Highly skilled and qualified in management, leadership, strategy and technology, Dehghanpisheh is currently the director of AI security at Palo Alto Networks, leading and hiring experts in AI and machine learning, business development, and digital marketing.