Foreign dominance in government AI access

The review of AI models provided for use by public officials found that foreign models — including those from Chinese developers — account for the majority of the offerings available. Among these are Chinese models that have come under US government sanctions, raising questions about the composition of the AI toolset available within the Korean public sector.

The finding highlights the reliance of Korean government systems on overseas large language models (LLMs) rather than domestically developed alternatives. No details on specific procurement figures or model vendors beyond this composition were included in the reviewed materials.

Security and procurement questions

The presence of US-sanctioned Chinese models among the tools accessible to officials is likely to intensify scrutiny of how AI services are vetted for government use. Oversight bodies such as the National Intelligence Service and the Ministry of Science and ICT have roles in shaping the security standards applied to AI adopted by the public sector, though the reviewed materials did not specify what screening, if any, was applied to the models in question.

The available evidence did not establish whether any of the foreign models are actively used in official work or merely listed as available options. That distinction remains unconfirmed.

The claim that a majority of AI models available to government officials are foreign-made, including Chinese models under US sanctions, is assessed as follows. Verdict: Mostly True