A free tool that stuns LLMs with thousands of invisible Unicode characters
18 hours ago
- #AI-plagiarism
- #Unicode-trick
- #text-obfuscation
- Tool inserts invisible zero-width Unicode characters between text characters to deter AI plagiarism.
- Effective for obfuscating up to 500 characters, especially key parts of essays.
- Some AI models crash, ignore, or respond confused to gibberified text.
- Useful for anti-plagiarism, LLM scraper obfuscation, or fun.
- Even a single gibberified word can disrupt tools like Flint AI.
- Tested against ChatGPT: fails to process or ignores the hidden characters.