Main Program

In this keynote, Jen Easterly looks at how AI arms defenders and attackers alike, and how to protect the enterprise without sacrificing innovation.

Drawing on her experience as the final victim of conman Hamish McLaren, Tracy Hall shows how trust is engineered and exploited, and how AI is now scaling it.

Frontier AI shortens the gap between a vulnerability being found and exploited, and the model put forward is governed agents on resilient infrastructure.

Real-world scenarios and gamification are two strategies Harry Ouaida uses to get cyber security students participating, shared at a Careers Village session.

Drawing on major incidents from 2020 to 2025, this talk sets out engineering principles and metrics for systems that withstand and recover from compromise.

Thirty years on, the password habits and social engineering flaws the film Hackers depicted remain unresolved, with lessons for teaching cyber awareness.

Stress, fatigue and cognitive overload shape security outcomes as much as technical controls, reframing human error as a systemic design problem.

Boards without cyber security expertise cannot properly assess emerging threats, and this session argues for bringing that perspective into governance.

Russia's Wagner Group has evolved from mercenary operations into disinformation and offensive cyber capability, raising fresh risks for Australian security.

Australian breaches increasingly trigger litigation, and building a defensible security program and contracts before an incident occurs limits legal exposure.

With AI-driven attacks reaching impact in as little as eleven minutes, this session sets out the architecture needed for autonomous, machine-speed SOC defence.

Building a SaaS platform security-first from day one sped up delivery and simplified operations far more than the team building it originally expected.

A pre-mortem exercise imagines a cyber incident has already occurred, then asks why the response failed, surfacing real gaps in decision-making and governance.

Applying cyber counterintelligence principles shifts detection from reactive indicators toward understanding adversary intent, targeting and likely adaptation.

The blueprint behind a working AI governance program at a major logistics provider, covering assurance, vendor risk, and what actually broke along the way.

The Model Context Protocol simplifies connecting AI agents to tools and data, but that same convenience opens dynamic, often unguarded runtime access.

A cricket umpire's decision review framework translates into a governance model for making AI outputs contestable, logged and defensible under audit.

Ethics statements and human-in-the-loop reviews are not real controls for autonomous AI agents; this session argues for hard runtime limits instead.

An interactive whodunit investigation puts attendees in the jury's seat to examine how gaps across people, process, and technology cause real breaches.

A developer with no security background used autonomous AI coding agents to place first in a CTF, revealing how agentic tools reshape attacker economics.