Tech Hub Stage

This session describes how AI has become mandatory in delivering machine speed cybersecurity defences today. But beyond the buzz and hype why are humans still critical in the loop?

Your SOC is a triage ward. Alerts arrive injured and humans patch them one by one. An immune system works differently. It learns, adapts and repairs while you sleep. Using Google SecOps, Wiz and agentic AI, this session shows how to build one, grounded in production numbers.

Supply chain cyber risks are no longer confined to your immediate third parties. They can hide deep in your extended supply chains, in 4th, 5th and nth parties. This session explores why visibility is the foundation of modern supply chain security, and how organisations can move from questionnaire-led, bilateral compliance activities to actionable intelligence and collaborative defence. With traditional approaches to TPRM struggling to keep pace with increasingly complex, interconnected supplier ecosystems, Haydn Brooks will explore how a collaborative, network-first approach can help organisations reduce risk at scale.

Modern software applications are 90% comprised of open-source dependencies, creating systemic fragilities traditional scanners miss. This session introduces the Cairn Index, a non-intrusive, metadata-driven approach to quantifying software supply chain resilience. Key takeaways include identifying "hidden" transitive risks and strategies for meeting regulatory mandates like DORA and SBOM compliance.

Autonomous and unmanned systems, aerial drones, ground vehicles, and robotic platforms now depend on AI for perception, navigation, and decision-making. That dependence creates an attack surface with no equivalent in conventional IT: adversarial inputs that survive the physical world, sensor spoofing, model extraction, and compromise of the model supply chain itself. This session examines how these attacks work against mission-critical platforms, why safety-critical and real-time constraints rule out many standard defences, and what actually holds up in deployment. Attendees will leave with a practical framework for assessing AI-related risk across their own autonomous systems.

IEC 63452 sets a new baseline for OT cybersecurity assurance in rail, against which new projects will be assessed. This session covers what the standard demands in practice, then shows how DTL uses CyRail, its AI-assisted compliance platform, to cut the evidence and assessment burden while keeping an auditable trail.

Vouchsec are doing the unthinkable - training fancy AI models to act like traditional computers. This might sound dumb but our bet is that this could be one of the most powerful tools for cyber defenders, and it's a bit harder than it sounds.

As cyber threats and AI-driven risks continue to evolve, organisations are rethinking traditional awareness and compliance programmes. This session explores the shift towards Secure Behaviour Management, examining how security teams can uncover hidden human risk, drive measurable behaviour change, and proactively reduce risk while demonstrating meaningful security and compliance outcomes.

Every security playbook assumes a human adversary. In this session, the human is obsolete. Watch live as a decentralized, open-weight AI swarm wakes up on stage. No operator. No off switch. No C2 server to burn down. You will see autonomous agents find each other, negotiate tasks, and execute a coordinated attack in real time. When the adversary is a self-healing hive thinking at machine speed, attribution is fiction, detection is a lagging indicator, and takedowns are impossible. Come watch the human era of security end.

SOCRadar’s Agentic Threat Intelligence platform empowers modern SOCs by scaling human judgment rather than replacing analysts. Discover how autonomous AI agents eliminate operational friction, automate workflows, and transform vast threat data into immediate, decisive action. 

Jake lifts the lid on AI's criminal underworld. Using face-swapping technology, he repeatedly passed live video job interviews under a false identity. Through real-world experiments, he reveals how criminals weaponise AI to clone voices, forge documents and infiltrate organisations in seconds. Nothing is real anymore. Key takeaways: • How AI face-swapping and voice cloning are being weaponised by criminals right now • Why deepfake candidates pose a genuine insider threat to recruitment and security • How deception can now be scaled in seconds with freely available tools

Cybersecurity is now a business leadership issue, affecting revenue, reputation, customer trust and resilience. This session explores the threats, regulatory scrutiny, supply chain complexity and stakeholder expectations shaping the next five years, offering practical insights to help boards and leaders prepare for, respond to and recover from disruption.

Over 99% of the UK's 25 million organisations are SMEs, most lacking dedicated cybersecurity resources. As key suppliers to government, critical infrastructure, & large businesses, they represent a major attack surface. This talk explores how to engage SMEs in cybersecurity and improve their resilience to strengthen the nation's overall cyber posture.

Redrawing the line between human and machine in offensive security: AI has quietly become both attacker and defender inside offensive security. Using real cases from the past year we explore where AI genuinely changes offensive security today, and where it does not. We'll look at AI for good, AI for bad, and how AI can be used for best in your cyber security toolkit.