Shay Weiss

This Think Tank explores the hidden risk side of GEO (Generative Engine Optimisation) and AI visibility. As organisations increasingly rely on LLMs, AI agents and AI-generated recommendations, the content these systems consume becomes a new enterprise risk surface. Through practical sandbox examples, the session will demonstrate how indirect prompt injection, poisoned context and misleading signals can influence AI answers, recommendations and downstream decisions, followed by a discussion on governance, ownership and mitigation strategies.

As LLM agents increasingly mediate shopping decisions, “agent visibility” becomes a new competitive front. This session explores how CDOs build structured content, schemas, trust signals, and guardrails that make brands easier for machines to understand—and recommend first. Join the panellists in a heated debate and examine how agent-ready data products create commercial uplift in a margin-pressured landscape.

• Understand what agent-to-agent commerce could mean for retail discovery.

• Discover how peers plan to test and measure their brand’s LLM visibility.

• Explore how panellists recommend to structure content, attributes, and proofs for agent parsing.

• Hear how to build teams, KPIs, and processes for ongoing agent optimisation.