The Ministry of Science and ICT said Wednesday its AI Safety Institute and OpenAI signed a memorandum of understanding to deepen cooperation on safety assessments in high-risk domains.
Under the pact, the two sides will share methodologies, benchmarks and best practices for evaluating high-risk AI applications. They will also exchange technical information to build assessment frameworks tailored to the Korean language and local social context.
The MOU broadens an October 2025 cooperation agreement between the ministry and OpenAI, extending it into the safety arena. Officials said the deal grew out of two rounds of talks this year between the ministry's second vice minister and senior OpenAI figures.
"As AI's impact on national core infrastructure and security grows, rigorous safety evaluation in high-risk areas is becoming not a choice but a necessity," said Kim Myung-joo, head of the AI Safety Institute. The institute, he added, would work with OpenAI to verify frontier-AI risks scientifically and help shape internationally accepted evaluation standards.
Lee Sang-hyun, OpenAI's Asia-Pacific policy lead, said Korea was a pivotal market where AI adoption and innovation were advancing at pace.
The institute and OpenAI plan to firm up specific tasks and timelines through working-level consultations, the ministry said.
Copyright ⓒ Aju Press All rights reserved.
