South Korea’s government has published a tender seeking private-sector partners to build a universal basic AI chatbot, a free-to-use service that would be available to every resident in the country.
The “AI for everyone” plan, administered by the Ministry of Science and ICT, calls for private entities to build and operate two AI systems under contracts expiring in 2031: a general-purpose chatbot available to all citizens, and an agentic system that lets people interact with government services through natural language.
Government offers GPU support
The tender documents reveal that Seoul will provide successful bidders with up to 256 Nvidia B200 GPUs, provided the winners match the government’s investment dollar-for-dollar. Bidders must also use locally developed AI models as the foundation for their services, a requirement designed to ensure data sovereignty and cultural alignment.
South Korean media report that local tech giants Kakao, Naver, SK Telecom, and LG are all expected to bid. The bid deadline is August 11, 2026.
Sovereign AI context
The initiative comes amid a global push for sovereign AI capabilities, accelerated by recent US government restrictions on Anthropic’s Mythos 5 and Fable 5 models. When the Trump administration compelled Anthropic to prohibit all foreign nationals from accessing the models, and Anthropic determined it could not implement passport checks, the company took both models offline entirely.
The incident spurred governments worldwide to reduce reliance on overseas AI providers. South Korea’s tender explicitly aims to ensure that its citizens cannot be cut off from AI services by policy decisions made abroad.
South Korea is well-positioned to execute the plan. The country has long nurtured domestic tech champions through protective policies, including restricting Google’s mapping services on national security grounds, which left Naver and Kakao’s mapping apps as the dominant alternatives.
Sources: South Korea to launch universal basic AI chatbot (The Register, July 15, 2026)