Citadel AI, Inc., a company that provides an automated quality control system for AI, has begun collaborating with Deloitte Tohmatsu Consulting LLC, a member of the Deloitte Tohmatsu Group, in the field of AI quality verification and monitoring.
With the advent of generative AI, the introduction of AI in companies is accelerating, and many solutions that allow easy use of AI are now available, advancing the “democratization of AI” so that anyone can utilize AI. Furthermore, this year is expected to see the spread of AI agents that autonomously make decisions and execute multiple tasks. AI is currently primarily used within the company, including corporate operations, but is expected to be used widely throughout the company, not only to improve productivity but also to create value.
On the other hand, while the risks of AI are beginning to be recognized, they have not yet been promoted in tandem with the use of AI, and companies need to be more proactive in focusing on AI risk countermeasures. Furthermore, while the number of target AI models is currently very small, it is expected that as the number of cases targeting AI use increases, this number will increase, and manual AI risk monitoring will reach its limits. It is essential to consider digitizing and automating risk monitoring as a more effective method.
In response to this situation, Deloitte Tohmatsu recognized the urgent need for a system that could provide comprehensive support for AI risk, from considering countermeasures to their implementation and operation, and therefore decided to collaborate with Citadel AI, which has AI quality verification technology.
こちらもお読みください： 日本経済新聞社がプロ向け生成AIサービス "NIKKEI KAI "を開始
デロイトトーマツ has been providing AI governance formulation and implementation support services using the “Trustworthy AI™” framework to help companies develop AI risk countermeasure policies and build and operate guidelines, processes, and systems. Through this collaboration, the two companies aim to help companies efficiently monitor AI risks while promoting the use of AI in a wide range of areas, thereby contributing to solving the issues companies face when using AI.
Through this collaboration, Citadel AI will provide advanced AI quality verification and monitoring technologies, including “Citadel Lens,” which automatically verifies AI models and data during training to accelerate quality improvements, “Citadel Radar,” which continuously monitors AI models and data during operation to support stable operation, and “Lens for LLMs,” a large-scale language model evaluation tool.
ソース PRタイムズ
