With the rapid adoption of generative AI, properly managing AI risks is becoming a growing concern for organizations, and AI risk management guidelines and regulations are being developed in many countries.
In this technical report, NTT DATA describes how they used Lens for LLMs, Citadel AI’s tool for evaluating LLM applications, to conduct an AI risk assessment of a production RAG (retrieval augmented generation) system. Their results demonstrated that Lens successfully detected multiple risks in this generative AI system.
This report can be read by not only engineers and managers involved in AI development, but for anyone interested in risk management of generative AI. For more details, please refer to the article on NTT DATA’s website, DATA INSIGHT, linked below (Japanese only).
https://www.nttdata.com/jp/ja/trends/data-insight/2024/1008/