Overview档案Profiles › Korea

韩国 · 企业 AI 高管教育与转型动员档案

调研截止:2026年6月24日
主要来源:三星/SK官方新闻稿、《서울경제》《이투데이》《Korea Herald》《Korea Times》、政府政策文件


一、三星集团:AX(AI Transformation)总动员

触发点与领导意志

2026年6月9日,三星集团(全体关联公司)联合宣布"AI大转换(AI 대전환)",由李在镕会长亲自指令:"把组织DNA彻底改变。"这被媒体类比为30年前李健熙的"除老婆孩子外全部换掉"——同等级别的顶层意志动员。

核心判断:三星认为,"CEO的AI文解力(AI literacy)决定AX的成败"——不是技术团队,是一把手必须先懂。

AX Boot Camp(社长团集训)

维度 内容
对象 全关联公司社长(사장단)约50名
时间 2026年6月,2天
地点 三星人力开发院·湖岩馆(삼성 인력개발원 호암관)
形式 实践型(非讲座型):高管亲手操作生成式AI,设计本司业务流程革新方案
产出 每位社长须当场发表《用AI重构本公司8大业务流程的方案》
8大流程 研发、采购、制造、物流、营销、销售、服务、经营支持

关键设计:社长不是被动听课,而是"交作业"——现场发表AI业务改造方案。将考核压力内化到教育环节。

全员AI教育滚动计划

层级 规模 时间节点
社长团(사장단) ~50人 2026年6月(已启动)
全体高管(임원) ~2,300人 截止2026年8月12日,分批2晚3天
全体员工 ~28万人 2026年底完成

AI工具:同步全面引入Gemini、ChatGPT、Claude三种外部生成式AI(2023年曾因泄密事故全面禁用,此次180度转向)。

组织侧配套

  • 全8大业务流程嵌入AI(非选修,是标配)
  • 目标:2030年前将工厂全面升级为AI自主工厂

二、SK集团:New Icheon Forum + AI Worker全员化

崔泰源会长的"全速前进"动员

2026年6月,SK集团首届New Icheon Forum(将战略会议与知州论坛合并)召开,崔泰源(Chey Tae-won)会长在场提出: - "现在是以全速、向全方位进入AI转型的时刻" - 提出"One Agent Per Person"——集团每名员工配备一个个人AI Agent - 强调:AI改变的不只是邮件质量,而是工作流程、汇报线、决策周期、产品规划的整体逻辑

SK Forum现场的AI实战

New Icheon Forum现场,AI Agent实时汇总高管讨论,并参与圆桌讨论(Panel)。换言之:AI不只是培训主题,而是当场成为"会议成员"。

SK Telecom:AI员工编号制

SK Telecom宣布为AI Agent分配员工编号(工号),赋予角色、职责与权限,与人类员工平行协作,建立对应的数据安全与权限治理结构。这是将AI纳入正式组织架构的最激进实验之一(韩国范围内)。

高管AI培训(SK Telecom / SK Broadband)

  • SK Telecom与SK Broadband对高管实施周末AI教育(待核实:具体课程内容尚未公开详细纲要)
  • 下阶段计划:以"AI for Customers"为主题,针对业务负责高管单独开展AI商业化策略培训
  • 工具覆盖:ChatGPT、Claude、Gemini、Microsoft Copilot

三、LG集团:制度化 > 运动式

구광모(Koo Kwang-mo)会长的AI人才战略

LG的路径与三星/SK不同:不强调急进式集训,而是建制度、立管道

项目 内容
LG AI大学院 2026年3月开办,韩国首家获教育部认可的企业内附属研究生院,招收员工就读硕士/博士
LG AI Academy 2021年起运营,面向员工的AI专业培训,逐步扩展为AI素养+专项项目领导力培训
LG Aimers 面向大学生,免费2个月AI实战课程+企业真实数据Hackathon,已运营至第8期
LG AI Youth Camp 与首尔大学联合,2晚3天线下+10周线上,前15名赴硅谷研修

评价:LG走的是"长线人才管道"路线,不做一次性轰炸,而是从大学生到博士生到在职员工分层培育。高管AI教育层面资料较少公开(待核实)。


四、现代汽车集团:Physical AI + 治理先行

郑义宣会长的框架

郑义宣(Chung Eui-sun)会长将AI定位为"不是一种技术,而是改变企业与产业'运作方式'本身的力量",强调:未能将AI能力内化的企业,无法保证生存。

方向差异:现代汽车的AI战略重心在Physical AI(自动驾驶、机器人、工厂),而非纯办公室流程自动化。

组织架构动作

  • 新设AI Governance TF,直属副会长张在勋
  • TF职能:AX PMO(项目统合)、AI风险应对、AX基础建设、AX项目总管、AX能力培育
  • 研究员AI培训:已对主要汽车企业研究员开展定制化GenAI实务课程(来源:现场讲师课后记录)

:相比三星/SK的"社长亲自上阵",现代汽车的高管层AI教育公开信息较少,执行细节待核实


五、韩国政府:举国AI人才政策

核心框架

政策 内容
AI基本法 2024年通过,2026年1月生效,全球第二个综合性AI法(仅次于欧盟AI Act)
AI人才培养全民计划 投入1.4万亿韩元(约9.6亿美元),覆盖小学→研究生院全阶段
5.5年快速通道 允许优秀学生在5.5年内完成本+硕+博,压缩传统8年学制
AI专业高中扩编 AI专项招生学校从14所翻倍至27所
劳动市场AI能力计划 雇佣劳动部推出"AI+能力提升项目",面向在职/转职人员强化AI使用能力

总统级表态:总统李在明将培育AI人才列为核心承诺,目标进入全球AI三强。

政策与企业行动的关系

政府提供人才管道与法律框架,财阀内部AI教育基本独立运营,政策协同以资金激励与人才供给为主,不强制要求企业具体培训形式(与中国工信部指令模式不同)。


六、横向比较与可借鉴点

各集团打法对比

维度 三星 SK LG 现代
领导人亲自动员 ✅ 李在镕指令+社长交作业 ✅ 崔泰源当场"全速进入" 🟡 会长更多表态,少现场 🟡 郑义宣框架表态
集训形式 AX Boot Camp 2天 New Icheon Forum 3天 无明显短期集训 无明显短期集训
制度化程度 中(运动式为主) 中高(结合论坛常态化) 高(研究生院+多层培训体系) 中(TF架构+待核实)
AI工具实战 现场操作+方案发表 Forum现场AI实时参与 员工Aimers+实际数据 研究员定制课程
全员AI计划 ✅ 28万人 2026底 ✅ One Agent Per Person ✅ 多层管道 待核实

对"一把手闭门日"设计的借鉴

借鉴点 三星AX Boot Camp的具体做法
不是听课,是交作业 社长现场发表"本公司AI流程改造方案"——将CEO变成执行者而非观众
领导意志前置 会长一句"改变DNA",全体系2个月内落地——靠的是韩国财阀结构,而非共识决策
层级分明的滚动动员 社长→高管→全员三波,不是一次性,是有节奏的压力传导
工具绑定 同步引入3种生成式AI工具,教育和工具普及同步走,避免教完没工具用
SK的论坛即实验 Forum现场AI实时运行——用Show don't tell让高管直接感受AI能力

反面教训(运动式风险)

风险 表现
动员与落地的断层 集训产出的"AI流程改造方案",回到日常后能否持续执行,目前无公开跟踪数据(待核实
形式合规 vs 实质变化 28万人2026年底完成教育——大规模铺开易流于签到式培训,内容深度存疑
韩国财阀结构的特殊性 "社长被点名上课"在财阀体系有强制力;在扁平化或中小企业无法复制
One Agent Per Person的落地难度 SK Telecom的AI员工编号制创意大,但治理框架与安全合规细节尚未公开(待核实

信息来源


Korea — Executive AI Education & Mobilization

Research cutoff: June 24, 2026
Primary sources: Samsung/SK official press, Seoul Economic Daily, Etoday, Korea Herald, Korea Times, government policy documents


1. Samsung Group: AX (AI Transformation) Full Mobilization

Trigger & Leadership Will

On June 9, 2026, Samsung Group (all affiliates) jointly declared an "AI Great Transformation (AI 대전환)." Chairman Lee Jae-yong personally issued the directive: "Change the organizational DNA entirely." Korean media compared this to Lee Kun-hee's 1993 "Change everything except your wife and children" — the same tier of top-down command mobilization.

Core belief: Samsung concluded that "CEO AI literacy determines whether AX succeeds or fails" — it is not the technology team but the CEO who must first internalize AI.

AX Boot Camp (President-Level Intensive)

Dimension Details
Target ~50 affiliate company presidents (사장단, the top executive tier)
Duration 2 days, June 2026
Venue Samsung Human Resources Development Institute, Hoam Hall
Format Practice-oriented (not lecture): executives operate generative AI tools directly and design actual business process innovation plans
Deliverable Each president must present on-site: "AI-based innovation plan for our company's 8 core business processes"
8 Processes R&D, procurement, manufacturing, logistics, marketing, sales, service, management support

Key design insight: Presidents do not passively receive training — they submit and present work. Performance pressure is embedded into the learning event itself.

Full-Company AI Education Rollout

Level Scale Timeline
Presidents (사장단) ~50 June 2026 (launched)
All executives (임원) ~2,300 By August 12, 2026 — batches of 2 nights / 3 days
All employees ~280,000 Completed by end of 2026

Tools: Samsung is simultaneously rolling out Gemini, ChatGPT, and Claude as officially approved external AI tools across all affiliates (a 180-degree reversal from the 2023 ChatGPT leak ban).

Organizational Complements

  • AI embedded across all 8 core business processes (not optional — a baseline requirement)
  • Target: all factories converted to AI-autonomous factories by 2030

2. SK Group: New Icheon Forum + AI Worker for All

Chairman Chey Tae-won's "Full Speed" Mobilization

In June 2026, SK Group held its inaugural New Icheon Forum (a merger of the annual strategy meeting and Icheon Forum), where Chairman Chey Tae-won: - Declared: "It is time to enter AI transformation at full speed, in all directions" - Proposed "One Agent Per Person" — every SK employee paired with a personal AI agent - Argued: AI changes not just email quality but workflows, reporting lines, decision cycles, and product planning

AI in Action at the Forum

At the New Icheon Forum itself, AI agents summarized executives' real-time discussions and participated in panel debates. AI was not only the topic of the event but also a live participant in it.

SK Telecom: Employee ID Numbers for AI Agents

SK Telecom announced it will assign employee identification numbers (employee IDs) to AI agents, granting roles, responsibilities, and permissions comparable to human employees, and establishing governance structures for data security and access protocols. This is among the most radical organizational moves to formally integrate AI into the corporate structure in Korea.

Executive AI Training (SK Telecom / SK Broadband)

  • SK Telecom and SK Broadband are running executive AI training on weekends (specific curriculum not yet publicly disclosed — unverified)
  • Next phase: separate AI training for business-line executives under the theme "AI for Customers," focused on developing AI-based commercialization strategies
  • Tools: ChatGPT, Claude, Gemini, Microsoft Copilot

3. LG Group: Institutionalization Over Mobilization

Chairman Koo Kwang-mo's AI Talent Pipeline Strategy

LG's approach diverges sharply from Samsung and SK: build institutions and pipelines, not campaigns.

Program Details
LG AI Graduate School Opened March 2026; Korea's first company-affiliated graduate school accredited by the Ministry of Education; admits employees for master's and doctoral degrees
LG AI Academy Running since 2021; expert AI training for employees, expanded to AI literacy and AI project leader tracks
LG Aimers Free 2-month AI practical course + hackathon using real LG company data for university students; currently in 8th cohort
LG AI Youth Camp 2-night/3-day offline camp at Seoul National University + 10-week online course; top 15 recipients study in Silicon Valley

Assessment: LG is pursuing a long-horizon talent pipeline rather than a one-time blitz. This produces deeper capability but slower short-term signaling. Executive-level AI training details are limited in public disclosures (unverified).


4. Hyundai Motor Group: Physical AI + Governance First

Chairman Chung Eui-sun's Framework

Chairman Chung Eui-sun has framed AI as "not merely a technology but a force that fundamentally changes how companies and industries operate," adding that firms unable to internalize AI capabilities cannot guarantee their survival.

Strategic differentiation: Hyundai's AI priority is Physical AI — autonomous driving, humanoid robotics, and factory automation — rather than pure office workflow automation.

Organizational Actions

  • Established AI Governance TF (Task Force), reporting directly to Vice Chairman Jang Jae-hun
  • TF functions: AX PMO (project integration), AI risk response, AX infrastructure, AX project oversight, AX capability development
  • Researcher-level AI training: customized GenAI practical courses delivered to researchers at major automotive subsidiaries (source: instructor post-session record)

Note: Compared to Samsung and SK, public information on executive-level AI training at Hyundai is limited. Implementation details are unverified.


5. Korean Government: National AI Mobilization Policy

Core Framework

Policy Content
AI Framework Act Passed 2024, effective January 2026; second comprehensive AI law globally after the EU AI Act
AI Talent for All Plan ₩1.4 trillion (~$960M) investment covering elementary school through postgraduate research
5.5-Year Fast Track Outstanding students complete bachelor's + master's + PhD in 5.5 years (vs. traditional 8+ years)
AI High School Expansion AI-specialized enrollment slots doubled from 14 to 27 institutions
Labor Market AI Competency Plan Ministry of Employment and Labor's "AI+ Competency Up Project" targeting workers entering, transitioning within, or returning to the labor market

Presidential commitment: President Lee Jae-myung named AI talent cultivation a core pledge, with a stated goal of becoming one of the world's top three AI powers.

Government vs. Corporate Relationship

Government provides talent pipeline and legal framework; chaebol AI education programs largely operate autonomously. Policy coordination is primarily through financial incentives and talent supply — not directive mandates on specific training formats (in contrast to China's MIIT-directive model).


6. Cross-Group Comparison & Lessons

Tactical Comparison

Dimension Samsung SK LG Hyundai
Chairman personally mobilizes ✅ Lee directive + presidents submit work ✅ Chey on stage "full speed" 🟡 Statement-level, less hands-on 🟡 Framework statement
Intensive short-form training AX Boot Camp, 2 days New Icheon Forum, 3 days None prominent None prominent
Institutionalization depth Low-medium (campaign-heavy) Medium-high (forum + normalized) High (graduate school + multi-tier) Medium (TF structure + unverified)
AI tool immersion Live operation + presentations AI agent at the forum itself Real company data in Aimers Researcher custom courses
All-employee AI plan ✅ 280K by end 2026 ✅ One Agent Per Person ✅ Multi-tier pipeline Unverified

Transferable Points for a "CEO Closed-Door Day" Design

Lesson Samsung/SK Concrete Practice
Output, not input Presidents must present AI process transformation plans on-site — CEO becomes actor, not audience
Leadership will as forcing function One sentence from the chairman → full rollout within 2 months — enabled by chaebol hierarchy, not consensus process
Layered wave rollout Presidents → executives → all employees, in timed batches with pressure transmission at each level
Tool co-deployment Three AI tools officially adopted simultaneously — no mismatch between "trained on" and "can actually use"
Show don't tell (SK) AI agent participates in the forum itself — executives experience capability in real-time, not via slides

Counter-Lessons (Campaign-Mode Risks)

Risk Manifestation
Mobilization-execution gap Presidents submitted on-site innovation plans — whether these are tracked and implemented post-bootcamp has no publicly available follow-up data (unverified)
Checkbox compliance Rolling out AI education to 280,000 employees by year-end creates pressure for attendance-driven rather than depth-driven training
Chaebol-specific structural leverage "Calling out presidents to attend class" carries coercive force in chaebol hierarchy; not replicable in flat organizations or SMEs
SK's AI Worker ambition vs. reality Assigning employee IDs to AI agents is conceptually bold, but governance, security, and liability frameworks are not yet publicly detailed (unverified)

Sources