Researcher Hung Nguyen discovered zero-day remote code exploits in Vim and GNU Emacs by utilizing Anthropic's Claude Code, a large language model. This finding highlights the double-edged nature of relying on AI tools for vulnerability detection, as they can rapidly identify security weaknesses but also introduce unforeseen risks. Nguyen's simple prompts to Claude Code uncovered the exploits, demonstrating the model's capability to efficiently identify vulnerabilities that might elude human developers. The fact that Claude Code was able to find zero-day exploits in widely used text editors like Vim and GNU Emacs raises concerns about the potential for similar vulnerabilities in other software. This discovery is particularly concerning given the rapid evolution of large language models like Claude Code1. As a result, developers and security teams must assess their exposure to these vulnerabilities and take immediate action to patch them, as the window for doing so is rapidly closing.