Science ministry opens public evaluation for sovereign AI model project

The science ministry said Tuesday it is recruiting 200 ordinary citizens to evaluate domestic artificial intelligence models as part of a project to develop sovereign foundational AI. The recruitment period runs from July 28 through August 4, with selected participants scheduled to test the systems between August 8 and August 11.

The user evaluation serves as a component of the second-stage assessment for the Independent Artificial Intelligence Foundation Model project, complementing benchmark and expert reviews. Kim Kyung-man, the ministry's director of AI policy, said the process allows actual users in the field to determine how useful the technology is.

Ryu Je-myung, the ministry's second vice minister, stated that the government has prioritized the development of independent AI models to resolve technological, cultural, economic and security dependencies on global models. He described the initiative as a core project aimed at developing models that can be freely used.

The evaluation will focus on four elite teams: LG AI Research, Upstage, SK Telecom and Motif Technologies. Ryu noted that the government plans to select one additional elite team to fill a vacant slot, offering opportunities to consortia that applied during the initial solicitation or those that did not pass the first stage but possess strong capabilities. This expansion follows a strategy to let more domestic firms access GPUs and participate in development.

Participants will be chosen based on quotas proportional to gender and age demographics. Interested individuals can apply through a QR code or via the website of the National IT Industry Promotion Agency (NIPA), which is co-managing the recruitment.

Detailed criteria and results for the project are expected to be released in August.