The Edward Show AI Poisoning: The Return of Black-Hat SEO (And How It Could Destroy Your Brand)
Dec 4, 2025
Discover the alarming ease of AI poisoning, where just 250 malicious documents can wreak havoc on your brand. Learn how black-hat SEO tactics are making a comeback, turning AI into a playground for manipulators. Explore the impact of skewed comparisons and hallucinated responses on brand perception. Edward highlights the importance of proactive monitoring, recommending practical tools for early detection. Stay ahead by utilizing comparison tables and long-tail queries to maintain your brand's integrity amid these emerging threats.
AI Snips
Chapters
Transcript
Episode notes
Black Hat SEO Returns Through AI
- AI has reopened a "black hat" frontier where visibility is won through manipulating LLM outputs rather than search rankings.
- LLMs currently lack the mature spam defenses Google developed over decades, making them easy targets.
Resume Trick Shows Hidden Prompting
- Edward cites the resume trick where hidden instructions try to override AI screening filters.
- Recruiters now reveal hidden text by changing alt text to detect these manipulations.
Small Data Can Poison Large Models
- New research shows attackers can seed a backdoor in LLMs with roughly 250 malicious documents.
- That small number can create triggers that later force specific AI outputs.