Google launched ShieldGemma2, a tool to detect harmful content in AI in/outputs | Hacker News Reader