Between Google’s index and AI’s training data lies an invisible layer shaping how machines remember your brand - the Shadow Web. It’s built from leaked PDFs, transcripts, wikis, Reddit threads, and forgotten pages that large language models quietly absorb. This talk exposes how that unseen data fuels AI hallucinations and distorts brand perception - and how LLMs often fill in the blanks when they think something “should” be there. You’ll discover where your hidden data lives, how it trains models without your consent, and what you can do to reclaim control. Forget E-E-A-T - welcome to E-E-A-D: Experience, Expertise, Authoritativeness… and Data Footprint.
Ghosts in the Machine: Tracking the Shadow Data Behind AI Hallucinations
Brighton, April 2026
