Earned Media Bias in AI Search

Earned media bias in AI search is the tendency of generative search engines to rely heavily on third-party editorial, review, comparison, institutional, and expert sources rather than brand-owned or social sources.

Key points

• The central empirical finding in the GEO paper is that AI search engines systematically favor earned media more than Google does across many query settings ^[src-028].
• In regional and vertical comparisons, Google preserved a more balanced mix of brand, social, and earned sources, while AI search frequently shifted toward publisher and review domains ^[src-028].
• ChatGPT and Claude were the most earned-heavy engines in several experiments, while Perplexity and Gemini included relatively more brand and social sources but still leaned earned overall ^[src-028].
• For banking prompts across personas, the paper found that engine identity explained more sourcing variation than the persona, with editorial financial media anchoring many answers ^[src-028].
• The strategic implication is that GEO cannot rely only on owned content. Brands need third-party reviews, expert mentions, authoritative roundups, and credible backlinks inside the sources AI engines already cite ^[src-028].

Related entities

• OpenAI
• Anthropic
• Gemini
• Perplexity API

Related concepts

• Generative Engine Optimization
• AI Search
• AI Search Citation Network
• Big Brand Bias in AI Search
• Machine-Scannable Content

Source references

• ^[src-028] Mahe Chen, Xiaoxuan Wang, Kaiwen Chen, Nick Koudas — “Generative Engine Optimization: How to Dominate AI Search” (2025-09-10)