Researchers have reverse-engineered how large language models like ChatGPT, Claude, Gemini, and Perplexity select which B2B websites to cite. This analysis reveals that these models prioritize content that is easily discoverable and structured for machine readability, even over traditional SEO factors. Additionally, a separate examination highlights significant security risks associated with exposing internal infrastructure to AI agents, particularly through protocols like the Model Context Protocol, which can create enterprise backdoors. AI
IMPACT Understanding AI citation patterns can help content creators optimize for AI discoverability, while awareness of infrastructure risks is crucial for enterprise AI security.
RANK_REASON The cluster consists of two separate analyses published on Mastodon, discussing how LLMs cite content and the security risks of AI agents, rather than a primary release or significant industry event.
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
