It’s been a rough month for generative AI (GenAI). First, AWS launched Amazon Q, its answer to Microsoft’s Copilot, only to have its own employees warn of “severe hallucinations” and data leaks. Then Google launched Gemini, its answer to ChatGPT, to much fanfare and an incredible demo, only to acknowledge after the fact that the demo was fake. Oh, and Meta released new open source tools for AI safety (hurray!) yet somehow failed to acknowledge the most egregiously unsafe aspect of GenAI tools: their susceptibility to prompt injection attacks.
To read this article in full, please click here
InfoWorld Cloud ComputingRead More