The rise of autonomous AI in cybersecurity

A new chapter in cybersecurity is unfolding. One where machines follow human-led commands but chart their own course. This shift is being driven by agentic AI: autonomous systems that can reason, plan and act independently to achieve complex goals. AI agents enable companies to rethink and reimagine the way they work.

In the context of cyber defence, these agents are not only more efficient tools but they’re also emerging collaborators. Imagine intelligent systems that detect threats in real time, coordinate responses across networks, probe for vulnerabilities and adjust their tactics as conditions change. They don’t wait for instructions. They act.

This is more than an upgrade in capability. It’s a shift in how cybersecurity is designed and led. And it raises big questions:

• How do we manage what we didn’t explicitly program?
• What does accountability look like when the actor is an algorithm?
• And most importantly, how do we lead in an era where intelligent agents will operate alongside, and sometimes, ahead of us?

This series from PwC’s Cyber & Risk Innovation Institute explores the new frontier of agentic AI in cybersecurity: the opportunities, threats and leadership needed to shape what comes next. Here’s a glimpse of what’s ahead:

Agentic AI isn’t just changing the tools we use; it’s rewriting the rules of engagement in cybersecurity. We’re stepping into a world where autonomous systems can defend, attack, adapt and evolve faster than humans ever could. For defenders, this is about more than keeping up. It’s about reimagining what’s possible.

The organisations that thrive in this new era won’t just adopt AI. They’ll build strategies, cultures and digital teams that think and act alongside us as teammates. This series is your map to that future. The age of agents has already begun. Let’s lead it.