Researchers discovered a security flaw in Google's Gemini AI chatbot that could put the 2 billion Gmail users in danger of being victims of an indirect prompt injection attack, which could lead to ...