Kakao, AI Safety Research Institute sign MOU to build AI safety evaluation framework

Kakao is teaming up with the AI Safety Research Institute to build a framework for evaluating AI safety, with the goal of establishing a system to identify AI risks and securing safety and reliability that meet global standards.

Kakao announced Tuesday that it signed an MOU with the AI Safety Research Institute on Monday to conduct joint AI safety assessments and build an evaluation framework. The two organizations plan to work together to establish an objective and systematic safety evaluation foundation covering AI models and AI agents.

Under the agreement, Kakao will provide its AI models and AI agents as subjects for safety evaluation, along with the infrastructure needed to carry out assessments and develop evaluation tools. The AI Safety Research Institute will be responsible for conducting evaluations and refining assessment methodologies. The two sides will jointly review evaluation results and collaborate on developing AI safety evaluation tools.

The partnership will proceed in three phases. The first phase will involve pre- and post-deployment safety assessments of language models. The scope will then expand to multimodal language models and AI agents. In the final phase, the two organizations will jointly develop evaluation tools capable of efficiently verifying the safety of AI models and agents.

Kakao said the agreement is significant in that it will build a system to proactively identify and verify risks in AI models while at the same time advancing the company's AI safety evaluation capabilities.

"We will continue to strive to deliver AI models and services that prioritize safety and trust alongside technological innovation," said Kim Se-ung, Kakao's AI Synergy Performance Leader.

"As AI technology grows more sophisticated, having a system that proactively identifies and assesses potential risks becomes more important than ever," said Kim Myung-ju, director of the AI Safety Research Institute. "We will do our utmost to build an AI safety evaluation framework that meets global standards."

Meanwhile, Kakao has been stepping up its efforts to strengthen AI safety. The company signed an MOU with KT Cloud to build a secure AI ecosystem and operates an advisory panel under its AI and Technology Ethics Subcommittee, composed of external experts.