The Security Chapter NA

AI-Powered Threats & the Human Perimeter

“AI has completely changed the economics of an attack. What used to take an adversary days now takes minutes — AI-assisted reconnaissance, automated exploit development, deepfake voice and video impersonation, and hyper-personalised phishing.

In 2026, the ‘human perimeter’ becomes even more exposed because attackers can convincingly imitate executives, colleagues, and even family members. The question is no
longer ‘Can someone fall for this?’ It’s now ‘How do we defend when AI can perfectly mimic trusted behaviour?’

This is why our first theme focuses on human-centric threats, AI-driven impersonation, and the controls that will actually move the needle.”

Enterprise AI Security & Guardrails for GenAI

“Every organisation is deploying copilots, building internal LLMs, and automating workflows through AI. But most are doing this faster than their governance can keep up.

The risks are real: prompt injection, data leakage, model tampering, ungoverned RAG systems, and AI agents that can take significant automated actions.

Combine that with the EU AI Act landing fully in 2026, and AI security becomes more than a technical challenge — it becomes regulatory, operational, and strategic.

This is why we’re dedicating time to secure-by-design AI, AI governance frameworks, and what a safe AI operating model actually looks like.”

Resilience Over Protection: The Realistic Future of Cyber Defence

“Protection will never be perfect, and 2026 is the year where regulators and boards start measuring organisations on recovery, not just defence.

Ransomware continues to evolve, attacks are designed to disrupt operations, and DORA raises the bar for resilience testing.

We’ll be digging into practical playbooks for:

• building resilience into architecture,
  
   
• running realistic crisis simulations,
  
   
• and proving to the board that the business can keep operating even under attack.”