Meta is building an encrypted chatbot after an AI agent went rogue and exposed sensitive user data. According to The Information, a Meta employee asked an AI agent a question on an internal forum, and the agent posted a response that led to massive data exposure for two hours. Meta is now working with Moxie Marlinspike (Signal creator) to bring end-to-end encryption to its AI chatbots. https://gizmodo.com/meta-is-building-an-encrypted-chatbot-after-ai-agents-went-rogue-and-expose-sensitive-data-2000735696 #AIagent #AI #GenAI #AIEthics #Meta