Someone Created First AI-Powered Ransomware Using OpenAI's gpt-oss:20b Model

ESET just dropped a bombshell: they've uncovered the first-ever AI-powered ransomware, dubbed PromptLock, and it's using OpenAI's gpt-oss:20b model to wreak havoc. This isn't your average malware—it's leveraging AI to generate variable Lua scripts on the fly, making detection a nightmare for security professionals.

The ransomware encrypts files and demands a ransom, but the twist is its use of prompt injection techniques to dynamically alter its behavior, evading traditional security measures. It's a game-changer in cyber threats, showing how AI can be weaponized for malicious purposes.

Researchers at ESET highlighted that this development complicates threat detection significantly, as the AI-generated scripts are highly variable and hard to pin down with signature-based defenses. It's a wake-up call for the cybersecurity community to ramp up AI-driven security solutions.

The article, published by Ravie Lakshmanan on August 27, 2025, underscores the urgent need for advanced AI countermeasures in the ongoing battle against evolving cyber threats. Stay vigilant, folks—this is just the beginning of AI-powered attacks.