MessiahGPT AI model facilitates automated ransomware and phishing attacks

Security researchers have identified MessiahGPT, a criminal artificial intelligence service being marketed on BreachForums to facilitate cyberattacks. The platform is designed to generate malicious software, including ransomware, phishing kits, rootkits, and credential stealers, without the safety restrictions or ethical guardrails found in mainstream AI models.

MessiahGPT operates on a freemium subscription model, offering free queries without registration and cryptocurrency-based subscriptions starting at approximately $8 per month. This low cost is intended to lower the barrier to entry for inexperienced cybercriminals. The service is accessible via a dedicated website and a Telegram community.

While the operators claim the model utilizes a Mixture-of-Experts design trained on dark-web archives and leaked documentation, Trellix researchers note that these technical specifications cannot be independently verified. Experts warn that such tools enable attackers to rapidly produce diverse variations of phishing emails and malicious scripts, potentially bypassing traditional security filters.