Kakao to Conduct Pre-Verification of AI Models and Agents Safety... MOU Signed with Institute for AI Safety

On September 29, Kakao announced that it has signed a strategic Memorandum of Understanding (MOU) with the Institute for AI Safety to jointly evaluate AI safety and establish an evaluation framework.

Through this MOU, the two organizations plan to expand their cooperation and build an objective, systematic safety evaluation foundation for artificial intelligence (AI) models and agents.

Kakao has signed a strategic business agreement (MOU) with the Institute for AI Safety to jointly evaluate AI safety and establish an evaluation system. At the signing ceremony held on the 28th, Kim Sewoong, AI Synergy Performance Leader at Kakao (left), and Kim Myungju, Director of the Institute for AI Safety, posed for a commemorative photo. Kakao announced this.

The reason Kakao signed the MOU is that the importance of AI safety has grown alongside the rapid advancement of AI technology. Last year, Kakao and the Institute for AI Safety jointly conducted a safety evaluation of Kakao's own language model, 'Kanana-1.5-9.8B.'

Under the agreement, Kakao will provide its AI models and agents that are subject to safety evaluation, as well as support the necessary infrastructure for conducting evaluations and developing evaluation tools. The Institute will be responsible for performing the evaluations and advancing methodologies. Both parties will jointly review the evaluation results and cooperate on the development of AI safety evaluation tools.

The collaboration between the two parties will proceed in three phases. Initially, they will conduct pre- and post-release safety evaluations of language models. The scope will then expand to include multimodal language models and AI agents. Finally, the two organizations will jointly develop evaluation tools that can efficiently verify the safety of AI models and agents.

Kim Sewoong, AI Synergy Performance Leader at Kakao, stated, "Through this collaboration, Kakao's AI models have established a more objective and rigorous safety verification system," adding, "We will continue to prioritize safety and trust as we deliver innovative AI models and services."

Kim Myungjoo, Director of the Institute for AI Safety, said, "As AI technology advances, systems that proactively identify and evaluate potential risks become even more important," and further noted, "We will do our best, through this partnership with Kakao, to establish an AI safety evaluation framework that meets global standards."

This content was produced with the assistance of AI translation services.