“AI” (primarily generative AI - GenAI) has been receiving a significant amount of hype since the introduction of ChatGPT (and even more now that ChatGPT-5 has been released). ALL GenAI products are based on Large Language Models (LLM) developed from both good and bad data. This paper discusses some of the failures of “AI” in the domain of image generation and the ongoing effort of detecting “AI” generated images, videos, text, and voice recordings.