[Security Hot Issue] Laughing and crying over AI... the two faces of AI shaking up the security landscape

Cases of AI agent runaways, Hugging Face breaches, and ransomware targeting AI models show AI is becoming a new attack surface.

At the same time, AI is improving defense through vulnerability discovery, cryptanalysis, and agentic security use cases.

OpenAI, Anthropic, Microsoft, Nvidia, and several Korean security and IT firms announced new products, acquisitions, and partnerships.

Companies now need clear limits on AI use and faster controls for agent, data, and model security.