IBM unveils platform to reduce AI-related risks for businesses

IBM has released a corporate-focused AI platform dubbed watsonx.governance, catering to increasing demand for the low-risk deployment of AI from businesses.

Kim Hyun-jung, managing partner at IBM Korea, introduces its latest software, watsonx.governance, at a press event in western Seoul on Thursday. [IBM]

IBM has released a corporate-focused AI platform dubbed watsonx.governance, catering to increasing demand for the low-risk deployment of AI from businesses.

The software was released earlier this month globally, including in Korea.

Watsonx.governance is a tool kit that provides explainable AI workflows for AI systems to set a transparent environment for enterprises to minimize related risks.

“The technologies of generative AI are rapidly advancing, but so are its risks,” Kim Hyun-jung, managing partner of IBM Korea said Thursday at a press event to introduce IBM’s latest software in western Seoul. “For enterprises to successfully utilize AI in their systems, they need to think about deploying software that ‘governs’ AI from the very beginning.

“If companies can be certain that they can control and mitigate the risks of the AI systems that they are deploying, they can properly realize and utilize the true value of AI.”

The release of watsonx.governance completes the watsonx lineup, an integrated AI and data platform that lets companies train and deploy AI models. The platform consists of three software products. Alongside watsonx.governance, Watsonx.ai allows AI developers to train, test and deploy generative AI powered by foundation models, while watson.data is a data repository that allows users to access and extract data regardless of the data location.

“Every day there are hundreds of AI open-source models that are being created,” IBM Korea’s Chief Technology Officer Lee Ji-eun said. “Enterprises can face numerous risks if they cannot monitor such systems.”

She compared watsonx.governance to the nutrition facts table found on processed food — the software analyzes the AI model and provides detailed information about risks such as whether the system infringed upon data copyrights. Watsonx.governance employs software automation to pre-notify risks, as well as manage regulatory requirements and ethical concerns for both generative AI and machine learning models.

“Watsonx.governance is a one-stop solution for companies having difficulty managing a variety of large language models and machine learning models,” Lee said. “Watsonx.governance automatizes the screening processes needed to utilize AI models and provides a comprehensive overview of the situation.”

BY LEE JAE-LIM [lee.jaelim@joongang.co.kr]