A federal judge is scrutinizing Google’s use of publisher content for AI Overviews, while new data reveals discrepancies between brand recommendations and actual website citation traffic.

Key Points

• Judge Amit Mehta questioned Google’s antitrust defense, suggesting that using publisher content for AI Overviews without control is unfair to content creators.
• Reports indicate Google’s August spam update targeted large-scale, automated content production rather than AI-generated material that provides genuine value.
• Google’s John Mueller stated that no specific Generative Engine Optimization (GEO) is required, as standard SEO fundamentals remain effective for AI search results.
• Ecommerce data shows that when AI tools recommend a brand, the brand’s own website receives the citation only 31% of the time.
• Testing on ChatGPT reveals that Reddit’s citation frequency is highly query-dependent, contradicting theories that the platform was broadly deprioritized in search results.

Why It Matters

These developments highlight a growing disconnect between brand visibility in AI answers and the actual traffic or referral value captured by websites. Businesses must now distinguish between being mentioned by an AI and securing the direct citations necessary to drive meaningful user engagement and conversions.