A new study has revealed a critical accessibility issue for content intended for modern AI assistants: popular crawlers, such as GPTBot, ClaudeBot, and PerplexityBot, do not execute JavaScript. This makes Single Page Application (SPA) based applications virtually invisible to neural network search, as they receive only an empty HTML skeleton instead of full content.
What Happened
An audit of 11 live websites revealed that when using Client-Side Rendering (CSR), crawlers receive only 5 to 8 words of text instead of thousands of words of actual content. Unlike Googlebot, which is capable of rendering JavaScript, the primary AI bots see only meta tags and the empty structures of React or Vue applications.
Context
Traditional search SEO and the new field of Generative Engine Optimization (GEO) differ architecturally. While classic search engines have adapted to script execution, the current generation of AI crawlers focuses on the rapid processing of raw HTML, creating a technological gap in data collection methodology.
Why It Matters for the Industry
For the industry, this signifies a shift from standard SEO to GEO. Developers need to implement Server-Side Rendering (SSR) using tools like Next.js or Astro, or utilize pre-rendering solutions. A website's technology stack now directly determines its representativeness in the responses of LLM-driven systems.
Why It Matters for Users
Website and brand owners who rely exclusively on Client-Side Rendering risk losing reach in ChatGPT Search or Perplexity responses. Their products and services simply fail to enter the context during RAG queries, leading to a decrease in organic traffic from new search channels.
What Is Not Yet Known / Limitations
There is a divergence in the assessment of consequences: ranging from technical debt in code to an existential risk for content distribution and business.
Sources
Author
Look at AI, Editorial Team