Dario Amodei in an interview
Anthropic / THE WORLD THEY ARE BUILDING

Powerful AI people can rely on.

A world where more people can turn what they know into work they could not finish alone.

See it taking shape ↓
01 / THE PRODUCTS

The products behind the promise.

Clawd pixel character in the official Code with Claude conference poster
Claude Enterprise interface with uploaded documents
Official September 2024 launch screenshot: adding documents to Claude. Historical UI, not a current-plan specification.
Claude Enterprise

Bring teams onto Claude with identity, access, administration and reporting controls.

Organizations can apply AI across real work while keeping enterprise controls around access and use.

Lotte Homeshopping

A partner gets guidance. QA gets time back.

30 to 40% fewer partner inquiries to QA staff

Moni uses Claude through Sendbird to interpret guidance and answer routine partner questions, leaving QA staff more capacity for complex cases.

What changed? ＋

Partners needed help interpreting test reports, documents and QA requirements. Repeated exchanges with specialists slowed the product launch process.

Customer-reported result in Anthropic’s case study; this is one implementation, not a general outcome guarantee.

Rakuten

A feature reaches the customer sooner.

24 days to 5: reported time to market

Teams use Claude Code for implementation work, giving engineers room to inspect changes and focus on the product they are delivering.

What changed? ＋

Engineering teams needed to change complex software without losing quality or security as they delivered new features.

Rakuten’s reported feature-delivery result in a company-published case study; workload and implementation affect outcomes.

Two people at a whiteboard with handwritten equations and diagrams; one points at the board and the other holds a marker.
A working session at a whiteboard, from Anthropic's company page.
02 / HOW IT BECOMES REAL

From a blocked task
to a result someone can use.

Follow a company adopting Claude. This is an illustrative enterprise journey, grounded in the product and the Korea CSM remit. Every useful deployment needs context, permissions, review and people who keep using it.

  1. 01

    Start with one task worth changing.

    Define the user, the input and the decision they need to make. A useful scope could be QA guidance, document analysis or a code change, with a business owner and a way to assess the result.

  2. 02

    The question needs the right context.

    Claude can work with the documents, code or connected information needed for a task. Enterprise identity and connector controls determine who can access which material.

  3. 03

    A useful result takes more than a fluent answer.

    An API application can embed Claude in an existing service. Claude Code works with a repository and development tools. Each surface needs instructions and access suited to the task.

  4. 04

    Someone must decide whether it is ready.

    Review the output against the customer’s requirements. Developers inspect changes and tests; business users validate the answer against their rules before relying on it.

  5. 05

    The pilot becomes a habit.

    Watch adoption and consumption, teach people how to use the workflow and record the value it creates. Expand when the customer can explain why another team should use it.

Follow the work, step by step

A safety lab that decided to sell

Anthropic calls itself an AI safety and research company. Its stated job is to build AI systems that are reliable, interpretable and steerable. The problem it starts from: AI is moving into real work faster than organisations can trust it. A bank, a telecom operator or a software team will hand a task to a model only if it behaves predictably, can be checked and stays inside the permissions it was given. The company is organised to match. It is a public benefit corporation whose stated purpose is the responsible development of advanced AI for the long-term benefit of humanity, and a Long-Term Benefit Trust helps elect its board alongside the stockholders.

a fundamentally positive future

Dario Amodei · Co-founder and CEO, Anthropic · Machines of Loving Grace, his essay · 2024-10

The belief that sets Anthropic apart is that caution and commerce point the same way. Dario Amodei has said the founders did not leave OpenAI because they disliked selling AI; he says he worked on commercialising GPT-3 himself. The disagreement was about how to do it. He calls the goal a race to the top: if one lab adopts a safety practice that works and rivals copy it, that counts as success. Daniela Amodei, the president, makes the commercial case: companies building daily workflows on AI want the model that hallucinates less, so publishing its limits works like a car maker publishing crash tests. In Korea the same logic shows up in a customer's own words: SK Telecom's AI collaboration lead said brand safety came first when it chose Claude.

Why the bet started to pay in 2025

Two things changed. The first is Claude Code, an agent that writes and changes software inside a developer's own tools, which became generally available in May 2025. The second is how fast enterprises bought once it worked. The company's run-rate revenue, its revenue pace scaled to a year, was $87 million at the start of 2024, about $1 billion at the start of 2025, over $5 billion in August 2025, $14 billion in February 2026 and above $47 billion in May 2026. In February 2026 the company said the figure had grown more than tenfold a year in each of the three previous years, and the chief executive links the jump to how customers now buy.

This is so productive.

Dario Amodei · Co-founder and CEO, Anthropic · Dwarkesh Podcast interview · 2026-02-13

What changes inside a customer

Follow one illustrative company. It picks one task worth changing, such as QA guidance for suppliers or a code migration, with a business owner who can judge the result. It connects documents, code and systems, with identity and connector controls deciding who may see what. Claude does the work through an API application, Claude Code in a repository or Claude Enterprise on employees' desks. People still inspect the output: developers review changes and tests, and business users check answers against their own rules. Then the pilot has to become a habit, with usage tracked, people trained and value written down. Amodei calls the middle steps fiddly: change management, security permissions and old software that must be rewritten before a model can be trusted inside it.

The safety story and the sales story meet at permissions and review. An economist who tracks business spending reads the rise more simply: Anthropic started with very technical customers, served them well and then broadened through tools such as Cowork. Both readings fit the numbers, and neither proves the other wrong. What a candidate can rely on is narrower: Anthropic sells to buyers who must answer for errors, and it sells trust as part of the product.

In Korea the fiddly middle is the job. Whichever of the 15 Seoul seats open on 8 October 2026 it is, most of the work is turning a capable model into a permitted, reviewed and adopted workflow inside a Korean organisation.

Understand the products and their customers

Three doors into the same model

Anthropic sells one family of models through several doors. Claude Enterprise is the door for a whole workforce. It gives employees secure access to chat, Claude Cowork, Claude Code and the company's own connectors, and it gives IT and security teams admin controls, management and visibility. The company says prompts, data and results are not used to train its models by default, a point that matters in the security reviews the Seoul sales postings describe. On 8 October 2026 the price page lists $20 per seat per month billed annually, plus usage at API rates, alongside a sales-assisted option. In July 2026 Samsung SDS said 20 Samsung affiliates, about 70,000 employees, were using Claude Enterprise.

Claude Code is the door for software teams. A developer hands Claude a bug fix, a test or a multi-day migration and steers and reviews it from the terminal, an IDE, Slack or the web. It can now run on a customer's own infrastructure in a public beta. Individuals get it in the Pro plan at $17 a month billed annually or $20 monthly, and Max plans start at $100 a month. It carries much of the company's growth: Claude Code's run-rate revenue was over $500 million in September 2025 and over $2.5 billion in February 2026, when enterprise use made up more than half of it and business subscriptions had quadrupled since the start of the year.

Claude Platform, the API, is the door for builders. A company calls the model from its own product or agent and pays for the tokens it uses; Claude Opus 5.5, released in September 2026, lists at $4 per million input tokens and $20 per million output tokens. Claude is also sold through Amazon Web Services, Google Cloud and Microsoft Azure, so a customer can buy it where its cloud contract already sits; SK Telecom runs it through Amazon Bedrock. Cowork lets knowledge workers delegate research, analysis and document tasks. A 16 September 2026 announcement began merging Cowork with chat into one Claude, starting on Pro and Max; Enterprise administrators receive at least 30 days of advance notice.

Claude is a little bit like an employee using Salesforce.

Dario Amodei · Co-founder and CEO, Anthropic · CNBC Closing Bell: Overtime, with Salesforce CEO Marc Benioff · 2026-08-26

Korean customers, in their own numbers

LG CNS, the IT services arm of LG Group, made Claude Code the engine of a service it sells, LG CNS Build Factory. On a construction company's 20-year-old project management system, it reports converting 2,888 of 2,913 APIs, a 99.1 percent completion rate, and moving 1,340 screens to React. The rebuild ran from November 2025 to June 2026 and, by LG CNS's own estimate, cost roughly half of a conventional rebuild. Its 200-engineer Build Center now uses Claude Code as its standard way to develop. The point LG CNS stresses is not speed: migrations its customers had deferred for years became plans they could execute.

SK Telecom, an investor since 2023, built call-centre tools on Claude: real-time document answers for agents during calls, and automatic summaries, topic tags, to-do lists and sentiment notes after them. It reports a 34 percent rise in agents' quality ratings of the model's answers during calls. Lotte Homeshopping uses Claude through Sendbird in Moni, an assistant that answers suppliers' QA questions before a product goes live; it reports 53 partner questions a day and 30 to 40 percent fewer inquiries reaching QA staff. Law&Company built SuperLawyer, a legal assistant, on Claude and reports 6,000 users within 180 days. When its usage surged, Anthropic's customer team helped it raise its API usage limits.

How to read these cases

Every number above is the customer's own result in a story Anthropic published, each about one engagement or one department. Treat them as what customers chose to measure, not as a benchmark. The pattern is still useful: a narrow first workflow, a measurable change, then expansion to more teams, as Lotte did when Moni grew from fashion suppliers to other product categories. That pattern is what the Seoul account and customer success roles are asked to repeat.

Pick one of these cases and be able to retell it in two minutes: the starting problem, what Claude changed, who measured it and which department would come next. Several Seoul postings ask for exactly this link between product adoption and revenue.

03 / THE EVIDENCE

The story, in numbers.

DATED OBSERVATIONS

Revenue run rate

USD · B
Revenue run rate2024: $87M; 2025: Approximately $1B; 2025-08: Over $5B; 2026-02-12: $14B; 2026-05-28: Over $47B$87MStart of 2024Approximately $1BStart of 2025Over $5BAug 2025$14BFeb 2026Over $47BMay 2026

Dated annualized revenue pace. Approximate values and lower thresholds retain the original wording; year-only dates follow the company’s approximate observation period. These are dated disclosures, not a continuous growth record.

Figures, dates & basis ＋
Revenue run rate
DateValueBasis
2024$87MAnthropic reported a run-rate observation of $87M at the start of 2024 in September 2025. The date follows that approximate period.
2025Approximately $1BApproximately $1B at the start of 2025, as the company dated it. This is the revenue pace at that point.
2025-08Over $5BExceeded $5B by August 2025, disclosed on September 2. The observation is dated to the reported month; 5 is the disclosed lower threshold.
2026-02-12$14BCompany-disclosed run rate as of the Series G announcement; not completed calendar-year 2026 revenue.
2026-05-28Over $47BCompany said run rate crossed $47B earlier in May. Date is the disclosure date; 47 is the lower threshold, not an exact annual total.

Anthropic

DATED OBSERVATIONS

Funding post-money valuation

USD · B
Funding post-money valuation2025-03-03: $61.5B; 2025-09-02: $183B; 2026-02-12: $380B; 2026-05-28: $965B$61.5BSeries E$183BSeries F$380BSeries G$965BSeries H

Company-disclosed funding transaction valuations. Do not add funding commitments to these valuations. The reported prospective IPO target is excluded because it is not a completed transaction. These are dated disclosures, not a continuous growth record.

Figures, dates & basis ＋
Funding post-money valuation
DateValueBasis
2025-03-03$61.5BPost-money valuation at this financing transaction; not the amount raised or a public market price.
2025-09-02$183BPost-money valuation at this financing transaction; not the amount raised or a public market price.
2026-02-12$380BPost-money valuation at this financing transaction; not the amount raised or a public market price.
2026-05-28$965BPost-money valuation at this financing transaction; not the amount raised or a public market price.

Anthropic

Claude Code moves into general availability.

The Claude 4 launch also makes Claude Code generally available, broadening the company’s route into software development.

Funding follows enterprise demand.

Series H funding supports research, compute capacity and the products and partnerships serving customers.

Implementation talent becomes a named investment.

Claude Frontier Academy announces a $100M commitment and a goal of training 10,000 engineers by the end of 2027. The number is a target, a future training goal.

Read the growth story

Seats, tokens and the clouds in between

Anthropic is paid in two basic ways. Organisations pay for access by the seat: individual Pro and Max plans, team plans and Claude Enterprise for a whole workforce. Builders pay for use: every call to the model through the API or a cloud platform is billed per token, the small unit of text a model reads or writes. The Seoul customer success posting names both, consumption-based and seat-based products, and asks the hire to fix underuse of either against what customers have committed to. Usage pricing cuts both ways for the customer. As teams do more, the bill rises: a May 2026 press analysis described AI coding costs outrunning budgets, including a technology chief who said his firm spent its whole 2026 coding-tools budget in four months.

Two revenue numbers circulate, and they answer different questions. Run-rate revenue is the company's annualised pace at one moment: above $47 billion in May 2026. Recognised revenue counts what was actually earned over a year: nearly $4.6 billion for 2025, up 12-fold, according to press reports of the IPO prospectus. The same reports put the 2025 net loss at $42 billion, of which roughly $34 billion was a non-cash accounting charge tied to earlier financing, and the operating loss at more than $8 billion. Compute and infrastructure cost $7.33 billion, more than half of $12.65 billion in operating expenses, and cash and short-term investments stood at $20.28 billion at the end of 2025. Customers are broad as well as large: over 300,000 business customers in September 2025, and more than 500 spending over $1 million on an annualized basis by February 2026.

The dated numbers and what each one measures
DateNumberWhat it measuresBasis
Start of 2024$87 millionRun-rate revenueCompany-disclosed
Start of 2025About $1 billionRun-rate revenueCompany-disclosed
Aug 2025Over $5 billionRun-rate revenueCompany-disclosed
Feb 2026$14 billionRun-rate revenueCompany-disclosed
May 2026Over $47 billionRun-rate revenueCompany-disclosed
FY2025Nearly $4.6 billionRecognised annual revenuePress-reported from the prospectus
Sep 2025Over 300,000Business customersCompany-disclosed
Feb 2026Over 500Customers spending over $1 million on an annualized basisCompany-disclosed
May 2026$965 billionSeries H post-money valuationCompany-disclosed
Q2 202653.3 percentClaude's share of paid generative AI card spending in KoreaResearch estimate

Where it stands against OpenAI

The main rival is OpenAI. Anthropic's Series H in May 2026 raised $65 billion at a $965 billion post-money valuation, above the $852 billion at which OpenAI was valued in late March; both are prices of private funding rounds. An expense-data index of more than 50,000 businesses found in May 2026 that 34.4 percent paid for Anthropic and 32.3 percent for OpenAI, up from 9 percent paying for Anthropic a year before. In Korea, card-payment data for April to June 2026 gave Claude 53.3 percent of paid generative AI spending, against 31.7 percent for ChatGPT and 11.7 percent for Gemini, with fewer but larger payments than ChatGPT. The first is a research estimate of paying-company share; the second estimates card-payment spending. Each has its own sample and denominator. The company itself claims the top share of enterprise AI.

if the growth rate is 5x a year instead of 10x a year, then you go bankrupt.

Dario Amodei · Co-founder and CEO, Anthropic · Dwarkesh Podcast interview, on buying compute ahead of demand · 2026-02-13

Three risks a seller should know

The first risk is that customers can leave. According to reporting on the prospectus, nearly a quarter of 2025 revenue came from two customers, and it warns that many of the largest clients are not locked into long-term contracts and could cut or stop spending. One case has been reported: Microsoft reportedly began cancelling most direct Claude Code licences in favour of GitHub Copilot CLI, while keeping a separate deal that offers Claude through its Foundry platform. The second is cost. Anthropic reports $518 billion of cloud, computing and infrastructure obligations over coming years, bought ahead of demand that has to keep compounding, which is the scenario the quote above describes. The third is politics. The Pentagon dispute and the June 2026 US export controls, which took the newest models offline for foreign users for about two and a half weeks, show that decisions in Washington reach customers in Seoul.

Two habits help in an interview. Say run-rate revenue with its date, and keep it apart from the nearly $4.6 billion of recognised 2025 revenue; Korean press often calls the run-rate ARR, which invites the comparison. And when you talk about growth in your segment, start with usage: in a consumption business revenue follows adoption, and adoption is what the Seoul postings ask their hires to drive.

04 / THE PEOPLE

Someone saw it first.

Official portrait of Dario Amodei
Co-founder & CEO

Dario Amodei

Dario Amodei co-founded Anthropic in 2021 with colleagues including his sister Daniela. He leads a company whose stated purpose is to make capable AI reliable, interpretable and steerable.

The choice behind the company ＋

The company puts safety and interpretability research alongside commercial deployment. That choice shapes what it studies, how it evaluates models and how it governs the business.

Official portrait of Daniela Amodei, co-founder and President
Co-founder & President

Daniela Amodei

Daniela Amodei is Anthropic’s co-founder and president. Her leadership complements the technical research organization with the work of building and operating a company.

The choice behind the company ＋

Anthropic describes an interdisciplinary organization across research, policy, product and operations. Selling enterprise AI requires those functions to work together around customer needs.

▶ Dario Amodei in conversation with Emily Chang.

“a fundamentally positive future”

Dario Amodei · Machines of Loving Grace, his essay · 2024-10
The decisions behind the company

Seven co-founders, one bet

Anthropic was founded in 2021 by Dario Amodei, his sister Daniela Amodei and colleagues. The company's leadership page lists seven co-founders: Dario Amodei, Daniela Amodei, Tom Brown, Jack Clark, Jared Kaplan, Sam McCandlish and Chris Olah. Dario Amodei is chief executive and leads the research direction and strategy. Daniela Amodei is president and chairs the board; she leads the work across research, engineering, product, governance and commercial execution and manages the executive team. Tom Brown, the Chief Compute Officer, leads the group that secures and scales the company's computing power, the resource every other plan depends on. By late 2025 the company had grown from 200 staff to more than 2,000 in a few years.

I believe we are entering a rite of passage, both turbulent and inevitable, which will test who we are as a species.

Dario Amodei · Co-founder and CEO, Anthropic · The Adolescence of Technology, his essay · 2026-01

Three decisions explain the company's shape. The first is governance: a public benefit corporation whose Long-Term Benefit Trust, with three trustees including Ben Bernanke, helps elect a board that includes Reed Hastings and Chris Liddell alongside the two Amodeis. The second is the customer. The international head told Korean reporters in June 2026 that Anthropic focused on safety and built its models for enterprises from the start. The third is holding lines under pressure. Korean press reports that the Pentagon designated Anthropic a supply chain risk in March 2026 after the company declined to lift its limits on mass surveillance in the United States and on fully autonomous weapons.

The people between Seoul and San Francisco

For a Korea hire the commercial chain matters more than the founders' list. Paul Smith, the Chief Commercial Officer, leads commercial strategy and global go-to-market operations. In September 2025 Chris Ciauri joined his international leadership group as Managing Director of International, working alongside heads for EMEA, Japan and the Americas. In May 2026 Anthropic named KiYoung Choi Representative Director of Korea. He joined from Snowflake, where he was general manager for Korea, after country leadership roles at Google Cloud, Adobe, Autodesk and Microsoft, over three decades in all. Krishna Rao, the Chief Financial Officer, leads financial strategy and operations and speaks for the company in its funding announcements.

Official portrait of Paul Smith, Anthropic's Chief Commercial Officer.
Paul Smith, Chief Commercial Officer, leads commercial strategy and global go-to-market operations.

How it says it works and hires

The company publishes seven values. Three say most about daily work: do the simple thing that works, which it explains as caring about the size of the impact rather than the sophistication of the method; a high-trust, low-ego organisation that communicates kindly and directly; and putting the mission first. On hiring, the careers page says Anthropic cares about what people can do, not where they learned it, and that about half its technical staff had no prior machine learning experience. For non-technical roles it looks for clarity, judgment and genuine interest in the mission, and it describes its policy, operations and business teams as small and high-impact, places where people shape how the company works. Non-technical interviews are conversational. Take-home tasks are done without Claude unless the company says otherwise, and staff are expected in an office at least a quarter of the time.

On the worry that headquarters decides and Seoul executes, the postings point both ways. Compute specialists are hired and managed centrally in San Francisco, the Revenue Strategy & Operations hire reports to an Asia-Pacific head in Sydney, global Revenue Operations owns the CRM, and in June 2026 policy work for Korea was handled by staff based in Tokyo. Yet the channel posting says much of the process for running partners in Korea has yet to be built, and that habits set with the first Korean partners will shape the programme in other countries. Authority sits with regional and global teams; much of the design work starts in Seoul.

05 / THE NEXT CHAPTER

Make Claude part of
how a Korean company works.

The local challenge is visible in a real Korean case: Lotte Homeshopping has already deployed Claude through Sendbird. Different enterprise customers need their own workflow, security review, rollout and value evidence.

LOCAL & REGIONAL LEADERSHIP

KiYoung Choi / 최기영

Representative Director, Korea (appointed May 2026)

LinkedIn ↗
Enterprise sales and partnerships

Enterprise GTM, cloud partnership, startup-platform and business-development functions appear in profiles observed with the Korea filter on 7 October 2026.

Applied AI and technical work

Applied AI and technical-staff headlines appear in the same observed profile set.

Korea-filter LinkedIn observation found 21 associated profiles on 7 October 2026, including community and academic affiliations. The groups summarize observed functions, not employee headcount or reporting relationships. The Korea leader is independently confirmed by the company and press.

Why Korea? Why now?＋

The sources give four reasons, and Signet reads them together. Users came first: in October 2025 Anthropic said Korea ranked in the top five for total and per-capita Claude use, that weekly active Claude Code users in Korea had grown sixfold in four months, and that a Korean engineer was the world's top Claude Code user; by May 2026 it said Koreans used Claude at more than 3.5 times the rate expected for the population. Money followed: Asia-Pacific run-rate revenue grew more than tenfold in the year to October 2025. Policy fits: the company says its Korea plans align with the national AI strategy, and its international head said Korea's AI Basic Act takes a risk-based approach that shares Anthropic's philosophy. Industry fits: Samsung and SK hynix joined the May 2026 funding as strategic infrastructure partners, and SK Telecom has invested since 2023. A May 2026 Korean trade report citing an industry source adds that Japan and India came first in Anthropic's plans and Korea moved up as interest grew.

SK Telecom invests

SK Telecom invests $100 million and forms a partnership to develop a model tailored to telecoms.

First Korea event

At Anthropic's first official event in Korea, the then Chief Product Officer says the year will be the first of full investment in Asia-Pacific and that a Korea branch is planned.

Korean entity set up

Anthropic Korea LLC is established, with an office in Gangnam.

First Korea hire

A head of APAC startup partnerships joins in Seoul from the Korean startup Coxwave.

Seoul office announced

Anthropic announces Seoul as its third Asia-Pacific office after Tokyo and Bengaluru, planned for early 2026.

Korea head named

KiYoung Choi, formerly Snowflake's general manager for Korea, becomes Representative Director of Korea.

Export controls hit the newest models

US export controls take Mythos 5 and Fable 5 offline for foreign users. The controls are lifted at the end of June.

Seoul office opens

At the launch press conference Anthropic names Naver, Nexon and LG CNS among Korean users and offers free Claude accounts to up to 60 researchers at the national AI research hub.

Samsung SDS partnership

Samsung SDS becomes a strategic partner; 20 Samsung affiliates and about 70,000 employees use Claude Enterprise.

Understand the Korea opportunity

Demand came before the office

Koreans were using Claude heavily before Anthropic had an office in Korea. In October 2025 the company said Korea ranked in the top five for both total and per-capita Claude use, that weekly active Claude Code users in Korea had grown sixfold in four months, and that a Korean software engineer was the top Claude Code user in the world. By May 2026 it said Koreans used Claude at more than 3.5 times the rate expected for the population. The office followed the users. A Korean entity was set up in July 2025, Seoul was announced in October 2025 as the third Asia-Pacific office after Tokyo and Bengaluru, and the office formally opened on 17 June 2026.

particularly for complex coding and enterprise applications

Paul Smith · Chief Commercial Officer, Anthropic · Announcement of the Seoul office · 2025-10-23

The company's reasons go beyond usage. It says its Korea focus will align with the national strategy to become an AI leader. Its international head told the launch press conference that Korea's AI Basic Act, in force since January 2026, takes a risk-based approach that shares Anthropic's philosophy. Samsung and SK hynix joined the Series H as strategic infrastructure partners, and SK Telecom has invested since 2023. Card-payment data shows Koreans paying, not only trying: Claude's monthly card payments in Korea rose from KRW 73.1 billion in April 2026 to KRW 95.3 billion in June, with corporate cards accounting for 60.9 percent of Claude’s second-quarter 2026 card payments. That is a research estimate of spending on Claude, not Anthropic's Korea revenue.

A Korea team built around partners

KiYoung Choi, Representative Director of Korea since May 2026, describes a partner-led plan. At the launch he said Anthropic was building a dedicated Seoul organisation, from technical engineers to operations and policy, and was reviewing infrastructure and data residency options for Korea. Hiring had started for technical, partner and demand-support staff; asked how many people Korea would have, he said the number in the organisation does not mean much. The first named Korea hire, in October 2025, was a head of APAC startup partnerships recruited from the Korean startup Coxwave. The partner track is already visible. LG CNS signed a Claude Enterprise contract in June 2026 and opened Claude to its employees, and Samsung SDS signed a strategic partnership in July to find new business together, train applied AI engineers and supply Claude Enterprise to Samsung affiliates.

creating real business impact matters

KiYoung Choi · Representative Director of Korea, Anthropic · Anthropic Korea launch press conference, Seoul, as reported in Korean; English rendering of the published Korean text · 2026-06-17

What the first hires are for

Read together, the 15 Seoul postings describe a full commercial unit being built at once. Five Enterprise Account Executives each own an industry (digital natives, financial services, healthcare and life sciences, manufacturing, retail and consumer goods) under a Sales Director, Enterprise who coaches them through Korea's conglomerate-centred buying. Three Applied AI roles prove Claude technically before and during a deal, one of them through system integrators and the AWS and Google Cloud partners. A Customer Success Manager, the first based in Korea, grows usage after the sale. A Channel Partner Manager builds the reseller and integrator channel. Revenue Strategy & Operations and the Finance & Strategy role for Korea run planning, territories and deal economics. Two seats sit outside sales: External Affairs, the first policy hire in Korea, and a Compute Country Lead whose posting speaks of bringing gigawatts of compute online in Korea.

Signet reads these postings as investment in several local functions: customer adoption, partner sales, policy engagement and compute development. The Compute Country Lead opening describes a future infrastructure build-out, not installed capacity. Local responsibilities also cross borders: the revenue-operations role reports to an Asia-Pacific head in Sydney, and compute specialists are managed centrally in San Francisco. The June 2026 export restrictions show that access to particular models can also depend on US policy.

Thirteen of the 15 postings mention both Korean and English, with different requirement strengths. The Sales Director requires business-level Korean and strongly prefers English; the partner architect lists both languages under preferred qualifications. Many roles ask you to build a territory, channel or process from its first steps. Signet reads that as a chance to shape a growing office, with the uncertainty and hands-on work that comes with it.

KOREA OPEN ROLES

15 open roles in Korea

Current Signet catalog · 2026-10-19

FOR YOU

Compare each role’s responsibilities and the strength of its conditions.

Solutions Engineer

Applied AI Architect

Seoul, South Korea

Posting checked

The work
Design customer evaluations and Claude integrations.
Who fits
Have customer-facing technical experience such as solutions architecture, sales engineering or technical account management.
Keep in mind
Support the Claude API and Claude for Work; travel occasionally for customer workshops.
Start a mock application ↗
Solutions Engineer

Applied AI Architects, Partner

Seoul, South Korea

Posting checked

The work
Enable integrators and cloud partners to build with Claude.
Who fits
A bachelor’s degree or an equivalent combination of education, training or experience in a field relevant to the role.
Keep in mind
Pre-sales partnerships with system integrators, AWS and GCP.
Start a mock application ↗
GTM · Sales · BD

Channel Partner Manager

Seoul, South Korea

Posting checked

The work
Build Korea reseller and integrator channels and partner deals.
Who fits
Have enterprise software or cloud channel/reseller relationship experience and experience with large Korean SIs or managed service providers.
Keep in mind
Own onboarding, deal registration, billing and local processes.
Start a mock application ↗
See all 15 roles
Operations

Compute Country Lead, Korea

Seoul, South Korea

Posting checked

The work
Coordinate Korea compute sites, power, financing and construction.
Who fits
Have an energy, infrastructure or related-sector background.
Keep in mind
Korean nationality and government/energy links strongly preferred.
Start a mock application ↗
Customer Success

Customer Success Manager

Seoul, South Korea

Posting checked

The work
Expand API and seat adoption as the first Korea-based CSM.
Who fits
Fluency in English and Korean is explicitly required, despite appearing under Preferred qualifications.
Keep in mind
Lead customer enablement, change management, value measurement and quarterly business reviews.
Start a mock application ↗
GTM · Sales · BD

Enterprise Account Executive - DNB

Seoul, South Korea

Posting checked

The work
Win new digital-native customers and build an outbound territory.
Who fits
A bachelor’s degree or an equivalent combination of education, training or experience in a field relevant to the role.
Keep in mind
Cover Korea's internet platforms, gaming, e-commerce, SaaS, fintech and media companies.
Start a mock application ↗
GTM · Sales · BD

Enterprise Account Executive - FSI

Seoul, South Korea

Posting checked

The work
Sell Claude across Korean financial services and compliance reviews.
Who fits
Direct experience navigating Financial Supervisory Service requirements and review processes is explicitly required.
Keep in mind
Win new accounts and expand existing relationships across banks, insurers, asset managers and fintech.
Start a mock application ↗
GTM · Sales · BD

Enterprise Account Executive - Healthcare & Life Science

Seoul, South Korea

Posting checked

The work
Develop healthcare and life-sciences enterprise customers.
Who fits
A bachelor’s degree or an equivalent combination of education, training or experience in a field relevant to the role.
Keep in mind
Navigate healthcare data, regulation and complex agreements.
Start a mock application ↗
GTM · Sales · BD

Enterprise Account Executive - Manufacturing

Seoul, South Korea

Posting checked

The work
Sell into Korean manufacturing groups with security and IP review.
Who fits
A bachelor’s degree or an equivalent combination of education, training or experience in a field relevant to the role.
Keep in mind
Navigate procurement and deployment across holding companies, affiliates and group IT service companies.
Start a mock application ↗
GTM · Sales · BD

Enterprise Account Executive - Retail/CPG

Seoul, South Korea

Posting checked

The work
Win and expand Korea retail and consumer-goods accounts.
Who fits
A bachelor’s degree or an equivalent combination of education, training or experience in a field relevant to the role.
Keep in mind
Coordinate seasonal buying cycles and multiple budget owners.
Start a mock application ↗
Marketing · Content

External Affairs, South Korea

Seoul, South Korea

Posting checked

The work
Build Korea policy relationships, submissions and AI-governance outreach.
Who fits
Have effective advocacy with Korean authorities and experience analyzing legislation and participating in regulatory processes.
Keep in mind
As Korea's first policy hire, build ties with government, researchers and civil society.
Start a mock application ↗
Finance · Legal

Finance & Strategy, GTM - Korea

Seoul, South Korea

Posting checked

The work
Model GTM investments, revenue goals, sales capacity and deal economics.
Who fits
Have experience across strategic/GTM finance, private/growth equity, VC, consulting or investment banking.
Keep in mind
Partner with sales, finance and accounting across functions.
Start a mock application ↗
Solutions Engineer

Manager, Applied AI Architect

Seoul, South Korea

Posting checked

The work
Lead pre-sales architects and guide Korean customer deployments.
Who fits
A bachelor’s degree or an equivalent combination of education, training or experience in a field relevant to the role.
Keep in mind
Design evaluations for Korean-language use cases; travel occasionally for customer workshops.
Start a mock application ↗
Strategy

Revenue Strategy & Operations

Seoul, South Korea

Posting checked

The work
Plan Korea targets, territories, forecasts and commercial operations.
Who fits
Have B2B SaaS/cloud/enterprise software sales strategy, revenue operations or commercial GTM experience.
Keep in mind
Initially no direct reports; report to the APAC head in Sydney, with occasional regional travel.
Start a mock application ↗
GTM · Sales · BD

Sales Director, Enterprise

Seoul, South Korea

Posting checked

The work
Coach Korea enterprise AEs and own the team’s revenue execution.
Who fits
Business-level Korean is explicitly required.
Keep in mind
Tailor sales strategy and AE training to enterprise procurement, budgets and decision processes.
Start a mock application ↗
07 / NOW & NEXT

The next scene is taking shape.

The next bottleneck is people who can build.

Anthropic announced Claude Frontier Academy, with a $100M commitment to develop enterprise implementation talent.

A model release changes the deployment equation.

Opus 5.5 adds capability and lowers published pricing relative to Opus 5. Anthropic’s performance and cost claims are company evaluations.

Capital and compute support the next scale.

The Series H announcement reports $65B of funding at a $965B post-money valuation, alongside a revenue run rate above $47B earlier in May.

More context on what comes next

An IPO with the risks in front

The past year ties back to the bet in three threads. The first is money in public view. Anthropic filed confidentially for an IPO in June 2026, and in late September press reports of the prospectus said about 80 of its 261 pages dealt with risks, against 48 on the business itself. The same coverage reported that the company is pursuing a $2 trillion valuation at listing, a press-reported target. For a company whose case rests on candour about risk, a prospectus that leads with risk is consistent.

A cheaper model, and a call to pace the frontier

The second thread is the product. Claude Opus 5.5, launched on 22 September 2026, performs at the level of Claude Fable 5.1 on most work, the company says, and costs 40 percent less to run than Opus 5 at default settings on typical workloads in company testing. It is the first release since Anthropic called for pacing the frontier, and outside evaluators including METR tested it before release. Ten days later the company launched Claude Frontier Academy, a $100 million commitment to train 10,000 Frontier Deployed Engineers by the end of 2027. Participation is by nomination, and those who pass lead a real Claude use case at their own organisation during a 12-week residency. The first cohorts run in San Francisco, New York and London.

we want those who graduate to set the standard for how AI gets built inside a business

Steve Corfield · Global Head of Business Development and Partnerships, Anthropic · Claude Frontier Academy announcement · 2026-10-02

What to watch next

The third thread is policy. On 25 September 2026 a US appeals court upheld, by two votes to one, the Pentagon's designation of Anthropic as a supply chain risk, and the company said it was considering all options, including further review. For a reader in Seoul, three things are worth watching over the next year.

  • Whether the listing happens, and how audited revenue compares with the run-rate the company announces.
  • Whether infrastructure and data residency options for Korea arrive, and whether Samsung SDS's applied AI engineer training reaches Korean customers.
  • Whether US policy again limits the models Korean customers can use, as the June 2026 export controls did.

A separate 16 September 2026 product change merges Cowork and chat into one Claude. It begins with Pro and Max over several weeks; Enterprise administrators receive at least 30 days of advance notice. Docs, Slides and Design are offered in beta on paid plans, with Enterprise administrators choosing when to enable them. This matters to a Korea account team: a feature announcement has to be translated into the customer’s plan, administrator choices and rollout schedule.

Their future. Your next chapter.

Find your part ↑

An independent profile by Signet. No affiliation or investment recommendation. Introductions happen only with your consent.
Research reviewed 2026-10-08

A LITTLE MORE CONTEXT

Understanding Anthropic.

What does Anthropic do?

A partner is waiting for a quality check. An engineer is staring at a codebase that needs changing. Anthropic puts Claude inside that work: understanding context, producing a useful next step and giving people something they can inspect.

What is the opportunity in Korea?

The local challenge is visible in a real Korean case: Lotte Homeshopping has already deployed Claude through Sendbird. Different enterprise customers need their own workflow, security review, rollout and value evidence.

Are these confirmed job openings?

Employer postings were checked on 2026-10-08. Availability can change. Mock applications request a review from Signet.

How can I explore the company in more depth?

Follow the company from its vision through products, evidence and people to the Korea team and current roles. Use the table of contents to move within the page, and expand the sections for more context.

Back to the collection ↗