GenAI risks and global adoption

In this episode of the Practical AI podcast, Daniel Whitenack and Chris Benson interview Rick Kobayashi and Kenny Song, co-founders of Citadel AI, about AI safety and security, particularly in Japan. The discussion covers Japan's unique position in AI adoption, balancing regulation with innovation, and the cultural acceptance of AI. The guests explain Citadel AI's role in helping organizations test, monitor, and govern their AI systems, focusing on mitigating risks like hallucinations and prompt injection. They delve into the importance of risk management, the balance between benefits and potential downsides of LLM implementation, and the necessity of both testing and continuous monitoring to ensure AI system reliability. The conversation also highlights Citadel AI's open-source tool, LangCheck, and explores the future of AI, emphasizing both the infinite opportunities and the need to address emerging risks.