Mandatory Release of Publicly Funded AI Data

The revised AI Basic Act, passed by the National Assembly, introduces a requirement that AI datasets created using government funds be made publicly available. Under the amendment, institutions and organizations that develop AI datasets with state support will be obliged to disclose them, rather than keeping them restricted.

The change is aimed at addressing one of the key bottlenecks facing the domestic AI industry: the scarcity of high-quality, Korean-language training data. By mandating public release, the government seeks to ensure that publicly funded data assets can be reused across the research community and by private-sector AI developers.

Implications for the AI Ecosystem

Wider availability of government-backed datasets is expected to lower the cost of developing AI models, particularly large language models (LLM) and domain-specific services that require substantial, localized training data. Researchers, startups and established developers stand to benefit from access to data that was previously locked behind institutional boundaries.

The revised act aligns with broader policy efforts led by bodies including the Ministry of Science and ICT to strengthen the domestic AI infrastructure, alongside ongoing initiatives such as government-backed AI dataset projects.

Details such as the exact timeline for implementation and the scope of datasets covered by the mandate were not specified in the available materials. The claim that the amendment has been passed and that public release of government-funded AI data will become mandatory is True.