PromptLock, the world's first ransomware leveraging artificial intelligence, has recently come to light. Leveraging OpenAI's open-source gpt-oss:20b language model, this software can autonomously generate malicious Lua code on compromised devices, displaying compatibility across Windows, Linux, and macOS systems. By employing pre-configured prompts, PromptLock generates code designed to search, steal, and encrypt files, demonstrating a high degree of stealth and cross-platform adaptability. While it currently lacks the capability to delete files, its potential threat cannot be underestimated. Attackers can further exploit external servers to run the model, circumventing local memory limitations. Experts caution that such attacks could signify a nascent trend in the malicious exploitation of AI, amidst the inadequacy of current defense mechanisms. OpenAI has assured that it has taken steps to mitigate the risk of model abuse and will persist in enhancing protective measures.