Naver joins hands with Saudi Arabia to develop Arabic LLM

Naver Corp., South Korea’s largest online platform, has signed a memorandum of understanding with the Saudi Data & AI Authority (SDAIA) to cooperate in creating an Arabic large language model (LLM), while jointly developing cloud computing and robotics technologies.

Naver said on Thursday it had signed the MOU on Tuesday on the sidelines of the Gain Summit 2024, a global AI summit hosted by Saudi Arabia, in Riyadh.

Under the preliminary agreement, Naver Cloud Corp. and Naver Labs Corp. will join hands with the state-run SDAIA to develop Arabic LLM and relevant services; build and manage data centers; offer cloud services; and develop AI robots and their applications.

SDAIA oversees AI and big data development. It is leading Saudi Vision 2030 projects under the leadership of Mohammed bin Salman, crown prince and prime minister of the Kingdom of Saudi Arabia​.

Earlier this year, Naver added Arabic translation to its AI-driven real-time translation service Papago.

Their preliminary agreement came after Naver and the Saudi Arabian government kicked off a five-year project in July to build digital twin platforms for Riyadh and four other major cities in the Middle Eastern country.

“In addition to our cooperation for digital twins, we will partner with Saudi Arabia's representative technology partner in the AI ​​field,” Naver said in a statement.

A digital twin is a virtual representation of a real-world physical product or system for practical purposes such as simulation, integration, testing and maintenance.

Naver pins high hopes on LLM, a new growth pillar for the internet giant. Last year, it debuted an upgraded version of its LLM HyperCLOVA X that boasts better Korean language skills than global sensation ChatGPT.

Saudi Arabia is leading the AI market in the Middle East. Its AI sector is forecast to reach $135.2 billion by 2030, making up 12.4% of the country's gross domestic product, according to PwC.

Write to Ji-Eun Jeong at jeong@hankyung.com

Yeonhee Kim edited this article.