Agenda

►The Challenge of Securing AI on a global scale

Yair Kler, Vice President, Security Architecture, DHL Group

• Learn how CISOs can enable responsible AI adoption by advancing innovation without resorting to a blanket “no” while preserving strong security foundations.
• Understand how FOMO‑driven AI adoption is pressuring enterprises into rapid, high‑risk decisions while bypassing established best practices.
• Addressing the issue of emerging AI‑specific threat classes such as prompt injection and why they create new risks that remain difficult to mitigate.
• Recognizing how long‑standing security challenges like secrets management and identity lifecycle governance are re‑emerging with greater complexity in AI‑driven environments.

►Beyond Human Identity: Securing AI Agents

Ioan Nascu, GenAI Security Assurance specialist, Citi

• Agentic Identity: Why AI agents require fundamentally different approaches to traditional Identity and Access Management (IAM)
• Emerging Threat Landscape: The unique risks that arise when intelligent agents become threat actors or targets
• Beyond Human Capacity: How oversight mechanisms work when AI agents outnumber, outpace, and outlast human administrators

►Designing Trusted AI: Secure-by-Default Architectures and AI-Enhanced SOC in Practice

Daniyal Naeem, Distinguished Engineer, Principal Security Authority - AI, BT

• How to design secure-by-default architectures for agentic AI systems, grounded in clear security policies and operational standards
• A practical MCP secure reference architecture
• Real-world use cases for AI-augmented SOC operations
• Key risk considerations when operationalising AI in security environments

►Invisible Leaks: The Hidden Risks of Chatting with AI

Manit Sahib, Ethical Hacker & Former Head of Penetration Testing & Red Teaming, Bank of England

• AI Privacy Risks: How tools like ChatGPT, Claude, and Co-Pilot can end up knowing more about you than your best friend (and never forget a thing). The hidden dangers of casually sharing information with AI
• When Small Details Add Up: Why a few “harmless” details can combine to paint a full picture & How scattered information can reveal sensitive data without you realising
• The Myth of Security: Why AI models aren’t as secure as we might think & How attackers can trick them into spilling information
• Simple, Practical Steps: For employees: how to keep personal and company data safe & For organisations: reducing AI-related risks before they grow

►Panel Discussion: Buying AI Without Buying Risk

Simon Brady, Event Moderator
Robert Cooper, IT Security Engineering Lead, easyjet
Ali Shepherd, Director of Cyber & Operational Resilience (CISO), FCA
Natalia Shevchuk, SVP, AI Security Architect, Citi
Jon Harrison, Tech Lead, Local AI Division, Ministry for Housing, Communities & Local Government

• When you’re buying an AI product, what’s the first security concern that comes to mind?
• When an AI vendor says their product is “secure,” what do you actually want to hear from them?
• What’s the fastest red flag that makes you pause or stop an AI purchase?
• What’s one question every procurement team should ask before signing an AI contract?
• If you could give one piece of advice to someone buying AI today, what would it be?

►Panel Discussion: Who Owns AI Risk? And How Do We Stay Compliant?

Simon Brady, Event Moderator
Jonathan Armstrong, Partner, Punter Southall Law 
Orlando Fernandez, Senior Technical Specialist at the Recovery, Resolution & Resilience team, Prudential Policy Directorate, Bank of England (BoE)
Adaora Ezennia, GRC Lead, THG PLC
Paul Jerram, Compliance and Responsible AI Officer, Keolis UK & Ireland

• In practice, who owns AI risk in your organisation — and is that ownership clearly defined at executive level?
• How are you ensuring AI use across the business stays aligned with existing regulatory obligations?
• What visibility do you have over third-party, embedded & shadow AI tools — and how does that impact your compliance posture?
• If asked to evidence AI governance to the board or a regulator tomorrow, how confident would you be?

►Humans are the weakest link? Think agAIn

Etay Maor, VP Threat Intelligence and Founding Member of Cato CTRL, Cato Networks

• Challenge the current approach to AI agents security
• Demonstrate risks and attacks against agents
• Look into the dark web and criminal forums as to what threat actors are saying