How South Korea plans to best OpenAI, Google, others with homegrown AI

From tech giants to startups, South Korean players are developing large language models tailored to their own language and culture, ready to compete with global heavyweights like OpenAI and Google.

Last month, the nation launched its most ambitious sovereign AI initiative to date, pledging ₩530 billion, (about $390 million), to five local companies building large-scale foundational models.

The move underscores Seoul’s desire to cut reliance on foreign AI technologies, hoping to strengthen national security and keep a tighter control over data in the AI era.

The organizations picked by the Ministry of Science and ICT to compete were LG AI Research, SK Telecom, Naver Cloud, NC AI, and the startup Upstage.

Every six months, the government will review the first cohort’s progress, cut underperformers, and continue funding the frontrunners until just two remain to lead the country’s sovereign AI drive.

Each player is bringing a different advantage to South Korea’s AI race. TechCrunch spoke with several of the selected companies about how they plan to take on OpenAI, Google, Anthropic and the rest on their home turf. NC AI declined to comment.

LG AI Research, the R&D unit of South Korean giant LG Group, offers Exaone 4.0, a hybrid reasoning AI model. The latest version blends broad language processing with the advanced reasoning features first introduced in the company’s earlier Exaone Deep model.

Exaone 4.0 (32B) already scores reasonably well against competitors on Artificial Analysis’s Intelligence Index benchmark (as does Upstage’s Solar Pro2). But it plans to improve and move up the ranks through its deep access to real-world industry data ranging from biotech to advanced materials and manufacturing.

It’s coupling that data with a focus on refining the data before feeding to the models to train. Instead of chasing sheer scale, LG wants to make the entire process more intelligent, so its AI can deliver real, practical value that goes beyond what general-purpose models can offer. “This is our fundamental approach,” co-head Honglak Lee told TechCrunch.

LG is improving its models via familiar tactics: offering them through APIs, then using the real-world data generated by users of those services to train the model to improve.

“As LG’s models improve, our partners can deliver better services, which in turn generate greater economic value and even richer data,” he said.

However, instead of chasing massive GPU clusters, LG AI Research is focusing on efficiency, getting the most out of every chip, and creating industry-specific models, he mentioned. The goal isn’t to outspend the global giants but to outsmart them with high-performing, yet more efficient, AI.