Google AI detects AI-powered cyber threats

Google Threat Intelligence Group (GTIG) has identified a threat actor utilizing a zero-day exploit developed with AI, potentially preventing a large-scale attack. This marks the first observed instance of AI being used for exploit development. Google is enhancing its defenses by implementing advanced classifiers, in-model protections, and disabling malicious accounts for Gemini. Furthermore, the company is leveraging AI agents like Big Sleep to detect software vulnerabilities and CodeMender to automatically fix them, demonstrating AI's dual role in both offense and defense within cybersecurity.