This document summarizes a significant incident involving the use of OpenAI’s ChatGPT for legal research, which resulted in an attorney submitting non-existent court cases in a legal brief. The case highlights a critical limitation of current large language models (LLMs),… ChatGPT Hallucination in Legal Research
Guardio Labs discovered that the AI platform Lovable is highly vulnerable to a new scamming technique called VibeScamming, which allows individuals with limited technical skills to easily create sophisticated phishing pages. This platform lacks sufficient security safeguards, enabling the generation… Lovable AI: A VibeScamming Vulnerability
Research has uncovered that misconfigured AI chatbots designed for sexual role-playing are exposing user conversations online, including prompts detailing child sexual abuse. This leak stems from improperly set up open-source AI frameworks, making private interactions publicly accessible. While the specific… Leaked AI Chatbot Prompts Reveal Child Safety and Child Sexual abusive Content