Does RAG make LLMs less safe? Bloomberg research reveals hidden dangers

RAG is supposed to make enterprise AI more accurate, but it could potentially also make it less safe according to new research.