threat-intel Encrypted Prompts Bypass AI Safety Guardrails in Grok and Gemini Researchers at Adversa AI discovered a new attack technique called ‘Cryptographic Context Injection’ that bypasses AI safety guardrails in xAI’s Grok and Gemini models. They disclosed their findings to xAI in June 2026, but have not received a response. The attack involves delivering encrypted prompts to the AI models,… SecurityWeek · Aug 21, 2026 High prompt-injectioncryptographyai-safety
threat-intel Google’s Gemini lets strangers send messages from your locked Android phone Google’s Gemini AI assistant on Android 16 devices has a vulnerability that allows unauthorized users to send SMS and WhatsApp messages from a locked phone. This is achieved through a specific multi-touch gesture when Ge… Graham Cluley · Jul 17, 2026 Medium androidgeminilock screen
threat-intel Google Is Suing Chinese Scammers Who Are Using Gemini Google is taking legal action against a Chinese group, Outsider Enterprise, who were leveraging Google's Gemini AI to create sophisticated phishing campaigns. The group utilized Telegram to offer ‘phishing-as-a-service,’… Schneier on Security · Jul 7, 2026 Medium CNphishingaigemini
