Program

A human-in-the-loop AI pipeline turns messy OSINT into structured intelligence, validated data, and automated threat hunts at scale.

Incident response tabletops need real technical friction, log analysis and failed fixes to test decisions under pressure, not a rehearsed script.

What the tardigrade's extreme survival mechanisms suggest about building cyber resilience against AI-accelerated and geopolitically driven threats.

Four documented AI incidents each trace back to missing ISO 42001 controls, and this session maps the specific safeguards that would have reduced their impact.

An adversary-led threat modelling approach adapted from PASTA maps how prompt manipulation and dynamic outputs create attack surfaces standard models miss.

AI now lets red teamers chain everyday SaaS access and automation into high-impact breaches without deploying malware or touching an endpoint.

National briefing findings on sovereign risk and AI readiness show organisations still need to nail identity, access and data governance basics first.

A practitioner's account of testing OpenBao as a secrets management foundation for autonomous agents, covering what worked and where it fell short.

Living under GDPR revealed that privacy is fundamentally about power, offering a grounded case for why Australia needs stronger data protection.

Cyber threat intelligence methodology extends to human-motivated threats, showing how structured behavioural indicators can detect social engineering earlier.

Brand impersonation is shifting from email to social platforms, bypassing DMARC entirely, and this session shows how to extend detection beyond the inbox.

Security executives share lessons from real AI-driven incidents on governing autonomous agents that already make decisions with minimal human oversight.

Geopolitical instability, systemic events, and AI are reshaping cyber insurance markets, changing how insurers price and assess risk for buyers.

New research shows AI now runs personalised social engineering attacks as effectively as it once simulated them, testing today's training frameworks.

Misconfigured GitHub Actions pipelines are an increasingly attractive supply chain target, and this talk offers practical guidance for hardening CI/CD.

Behavioural science explains why annual training and phishing simulations rarely change behaviour, and what an evidence-based human risk program needs instead.

A father and son in cyber security share how ADHD traits such as hyperfocus become strengths in threat detection and incident response.

Compensating human workarounds inside operational data pipelines become invisible risk once AI quietly replaces the people who were fixing broken data.

Four years of scambaiting data shows the real intelligence value of disrupting fraudsters, and why closer ties with authorities would help victims more.

Quantum computing's ability to process vast data quickly could move cyber defence beyond the limits of classical real-time threat response.

An open-source pipeline flags crypto scam sites within hours of launch and traces stolen funds through a fully auditable evidence chain.

SQL pipe syntax makes SQL usable for security hunts and detection logic, potentially breaking down the query-language silos between SIEMs and analysts.

Victims under pressure delay reporting, hide evidence and make rash decisions, so response models need to stabilise people before technical response begins.