London · South Place Hotel
Hosted by Corinium Intelligence
A private London dinner for senior data, AI, technology and security leaders exploring secure runtime data controls for enterprise AI agents on 13 October 2026. Hosted by Skyflow and organised by Corinium Intelligence, The Secure AI Paradigm brings a select group of 10–15 executives to South Place Hotel in London. The focus is a critical production challenge: giving autonomous AI systems the information they need while retaining control over how sensitive data is accessed, processed and revealed. What the roundtable will explore As AI agents begin retrieving information, calling tools and taking action independently, static permissions and conventional security boundaries may no longer provide sufficient protection. This closed-door conversation will examine practical ways to embed privacy, governance and least-privilege access directly into enterprise AI architecture. Discussion points include: - Deciding which sensitive data an AI agent genuinely needs - Applying dynamic, policy-driven controls based on identity, agent, task and context - Protecting data as it moves through RAG systems, models, APIs and tools - Managing agent-to-tool and agent-to-agent connections - Protecting PII without stripping away the context AI needs to perform effectively - Translating security, privacy and regulatory requirements into enforceable technical controls - Identifying where existing access controls fall short for autonomous workflows The group will also consider what must change when an agent can independently retrieve information and act, and what safeguards would give enterprises confidence to scale agents across sensitive workflows. Roundtable hosts The discussion will be hosted by Javier Morao, EMEA Head of Data & AI at Citi, who leads data commercialisation across more than 50 countries and has over 15 years of experience in data, analytics and transformation within global financial services. He is joined by Roshmik Saha, Co-founder and CTO of Skyflow, a technologist with more than 15 years of experience building secure, scalable platforms. His background spans Microsoft and leadership of Lyft’s self-driving car infrastructure. View the event and registration details
