Gibberifier

Gibberifier is a web-based tool designed to prevent large language models (LLMs) and AI-powered web scrapers from extracting or interpreting text. It addresses the problem of automated systems reading and processing content by inserting thousands of invisible zero-width Unicode characters between each character of the input text. This transformation makes the text appear unchanged to human readers but disrupts the ability of LLMs and web scrapers to tokenize and understand the content.

The tool allows users to adjust the intensity of the gibberification process and provides immediate output, showing both the gibberified text and how AI models perceive it as a sequence of seemingly random Unicode byte tokens. Gibberifier is intended for use cases such as anti-plagiarism, obfuscating text to block LLM scrapers, or simply for entertainment. It is particularly effective when applied to important sections of text, up to about 500 characters, and is compatible with platforms like Google Docs.

Gibberifier has been tested against several AI models, including ChatGPT, Claude, Gemini, Meta AI, Grok, and Perplexity AI, as well as AI-powered web scraping tools like Firecrawl and TLDR This. In these tests, the gibberified text caused models to either crash, fail to process, or return confused or meaningless responses. The tool is suitable for individuals seeking to protect their written content from automated extraction or for those interested in experimenting with text obfuscation techniques.

Gibberifier is delivered as an online service accessible via its website. It functions as a specialized tool for text obfuscation and AI-blocking within the broader class of infrastructure solutions aimed at content protection.