Hackers Used AI to Develop First Known Zero-Day 2FA Bypass for Mass Exploitation

Google has identified the first known zero-day vulnerability exploited in the wild, which was likely developed using an AI system, marking a significant escalation in malicious AI usage for vulnerability discovery and exploit generation.

• •The zero-day targets a 2FA bypass vulnerability in a popular open-source system administration tool, implemented through a Python script exhibiting hallmarks of LLM-generated code with educational docstrings and textbook formatting.
• •Multiple threat actors including UNC2814, APT45, APT27, and TeamPCP are weaponizing AI models like Gemini to accelerate vulnerability discovery, malware development, and autonomous exploit creation with minimal human oversight.
• •PromptSpy Android malware abuses Gemini for autonomous UI navigation, biometric data capture for authentication bypass, and dynamic C2 infrastructure updates without requiring payload redeployment.