Why "Anthropic" Is Trending in South Korea: Samsung, Naver, Nexon and a November IPO

Photo by Igor Omilaev on Unsplash

If you follow Korean search trends the way you follow comeback schedules or drama ratings, you may have noticed something unusual sitting between the food fads and the sports headlines this week: 앤트로픽 — the Korean transliteration of Anthropic, the American AI lab behind Claude.

That's a strange thing to see trending in Korea. Korea has its own large language models, its own platform giants, and a government that talks openly about "sovereign AI." So why is a San Francisco company suddenly a Korean search term? The short answer: Anthropic has quietly become the coding and knowledge-work layer inside some of Korea's biggest companies — Samsung SDS, Naver, LG CNS, Nexon — right as it marches toward a very large IPO.

Two stories colliding at once

1. The IPO march

Korean business media has been running fresh coverage of Anthropic's path to a public listing. Reports in the Korean press, citing wire services such as Bloomberg, describe the company expanding its revolving credit facility significantly ahead of a listing, with a cluster of major Wall Street banks — Morgan Stanley, Goldman Sachs, JPMorgan and Citigroup — named as underwriters, and a November timeline floated. Valuation figures circulating in coverage are enormous, and Korean outlets have framed the story as Anthropic possibly reaching the public market ahead of OpenAI.

A caveat worth stating plainly: much of this is press reporting sourced from wires, not confirmed regulatory filings. Timelines slip, valuations move, and underwriter lineups change. Treat the numbers you see recirculating on social media as directional, not final. If you care about the specifics, go to the primary sources — company newsroom posts and filings — rather than aggregated screenshots.

2. The Korea business footprint

The second, more durable story is local. Anthropic said in late 2025 that it would open a Korean office in Gangnam, Seoul, and that office formally opened on June 17, 2026. It's the company's third major Asia-Pacific base, after Tokyo and Bengaluru. The opening came bundled with a set of enterprise, research and non-profit partnerships across the Korean AI ecosystem.

Which Korean companies actually use Claude?

• Samsung SDS — the Samsung Group's IT services arm — is deploying Claude, including Claude Code and Claude Cowork, to Samsung Electronics employees for knowledge work, agentic workflows and software development.
• Naver has rolled out Claude Code across its engineering organization, reportedly reaching thousands of engineers.
• LG CNS is extending Claude to thousands of employees for software work.
• Nexon, one of Korea's largest game publishers, is adopting Claude — the detail that should make gamers sit up.
• Wrtn Technologies, a Korean AI service startup, has a multi-year relationship with Anthropic.

Why this matters more than the deal count suggests

Korea is one of Asia's densest enterprise-software markets, but it's also one of the most concentrated. A relatively small group of conglomerates — Samsung, LG, SK, Hyundai, Naver — accounts for an outsized share of national software and IT spending. Winning Samsung SDS and Naver isn't "two logos on a slide." It's access to two of the largest engineering organizations in the country, each with downstream suppliers, subsidiaries and partners that tend to standardize on whatever the parent company uses.

There's also a business-model dimension. Korean coverage repeatedly notes that roughly 80% of Anthropic's revenue comes from enterprise customers, with Claude Code driving growth in corporate development work. That's a meaningfully different posture from consumer-chatbot competition. Anthropic isn't trying to become Korea's national model — it's trying to become the tool developers open every morning.

The competitive picture: foreign models as a layer, not a replacement

This is the nuance most English-language coverage misses. Korea has homegrown foundation models — Naver's HyperCLOVA X and LG's EXAONE — and active government policy pushing sovereign AI capability. So foreign models generally get adopted as productivity and coding layers rather than as the strategic core. Naver can deploy Claude Code to its engineers while continuing to build HyperCLOVA X; those are not contradictory decisions. One is a developer tool, the other is a national-scale platform bet.

For readers tracking the "Anthropic vs. OpenAI in Asia" storyline, the office-expansion race is the clearest scoreboard: Tokyo, Bengaluru, Seoul. Anthropic's positioning in Korea has leaned enterprise-first and coding-first rather than consumer-app-first, which is a distinct strategy from courting mass users.

What this means for Korean developers — and gamers

If you play Korean games, the Nexon detail is the one to watch. AI-assisted coding inside a major publisher's engineering pipeline doesn't change what a game feels like overnight, but over a few years it tends to show up as faster patch cycles, more localization throughput, and more tooling built internally instead of licensed. The same logic applies to webtoon platforms and content studios that sit under the same corporate umbrellas.

For developers: Claude Code being standard issue at Naver and LG CNS meaningfully changes what a Korean engineering résumé looks like. Familiarity with agentic coding workflows is drifting from "nice to have" toward baseline expectation at large Korean tech employers.

The skeptical section (because it's needed)

A balanced read requires acknowledging the friction. Korean reports covering Anthropic have also referenced the company's usage-policy disputes with the US government regarding surveillance and autonomous-weapons applications, as well as a federal court ruling that went in Anthropic's favor. Those two facts pull in different directions and are worth understanding rather than cherry-picking.

Other things to keep in perspective:

• Pilots aren't permanence. Large Korean enterprises run many vendor trials simultaneously. Deployment announcements are real, but multi-year lock-in is a separate question.
• Policy risk. Sovereign-AI initiatives could shift procurement preferences toward domestic models in public-sector and regulated contexts.
• IPO numbers are provisional. As noted above, verify against primary sources.

A small naming note

If you're searching Korean sources, you'll see the company written both as 앤트로픽 and 앤스로픽. Both refer to Anthropic; 앤트로픽 is the spelling currently trending. Searching only one form will hide half the coverage.

FAQ

Is Anthropic going public, and when?

Reporting in Korean and international press points to a November timeline with a major-bank underwriter lineup, but this is press reporting rather than confirmed filing detail. Check official filings and the company newsroom before treating any date or valuation as settled.

Which Korean companies use Claude?

Publicly reported adopters include Samsung SDS (serving Samsung Electronics employees), Naver, LG CNS, the game publisher Nexon, and the AI startup Wrtn Technologies.

Does this mean Korea is abandoning its own AI models?

No. Naver's HyperCLOVA X and LG's EXAONE remain active strategic programs. Foreign models like Claude are typically layered in for coding and productivity work, not adopted as replacements for domestic foundation-model efforts.