If AI is cutting entry-level work, why are LG CNS and Samsung SDS still hiring graduates?

The same AI shock is producing opposite hiring scenes

In Korea's platform sector, large-scale graduate hiring is being recalculated. NAVER had not finalized a 2026 mass-recruitment plan while reviewing talent profiles and hiring methods for the AI era. Kakao faced a separate organizational restructuring context and had not committed to repeating the previous year's group-wide intake.

Korean IT services firms show the opposite scene. Reporting on September 23 identified Samsung SDS, LG CNS, POSCO DX, CJ OliveNetworks, Hyundai AutoEver, Lotte Innovate and Shinsegae I&C as running second-half graduate recruitment.

The idea that AI simply removes junior work is not enough to explain both scenes. The more useful question is where each company sits in the AI value chain.

LG CNS is hiring a triple-digit graduate cohort across nine job families

LG CNS recruited graduates across AI, robotics, consulting, DX engineering, cloud application modernization, architecture, ERP, smart factory and convergence engineering.

Its official description defines the AI role around designing and implementing AI models and services for client businesses. The robotics role applies robots in manufacturing and logistics environments and tests real-world deployment feasibility.

For this kind of company, AI adoption is not only a force that can reduce internal workload. It is also customer demand that creates delivery projects.

CJ OliveNetworks changed both whether it hires graduates and what it expects from them

CJ OliveNetworks opened graduate recruitment across 11 fields including AI, data, software, DX, ERP, information security and cloud.

At the same time, it introduced AI capability assessment for every role. Development applicants face a vibe-coding test, while non-development applicants are evaluated on using AI in work scenarios.

This is not simply more graduate hiring. The company is continuing to hire graduates while raising the starting baseline toward people who can already work with AI.

For IT services firms, AI is both a productivity tool and a project they can sell

At a platform company, AI can raise productivity in search, development, document work and operations, compressing parts of existing workload.

IT services firms face an additional effect. As clients and group affiliates adopt AI, more projects are needed to connect models to existing systems, data, cloud infrastructure and business processes.

Recent Korean industry reporting points to growing AX programs and client projects as a driver of demand for delivery talent, including graduates. The same technology can therefore create efficiency pressure in one organization and new implementation demand in another.

If AI changes early-career productivity, it also changes hiring economics

Industry interviews point to another shift: generative AI may be improving the productivity of new employees earlier in their tenure. Work that once required a longer ramp-up can sometimes be completed sooner when AI is used as a working tool.

This is not a quantitative finding that every graduate's productivity has risen by the same amount. It is safer to treat it as an industry observation that is directionally consistent with changing assessment methods.

Still, the economics can change. When experienced AI and cloud talent is scarce and expensive while project demand is growing, hiring graduates who can become productive faster with AI and developing them internally can become more attractive.

What may be disappearing is not graduates, but the old starting point for graduates

LG CNS assesses technology and industry awareness across roles and uses coding- or planning-oriented formats to assess generative-AI capability by job. CJ OliveNetworks also separates AI-use assessment for development and non-development roles.

The detailed tests differ, but the direction is similar. Being entry-level no longer automatically means arriving with no evidence of using AI in real work.

Rather than choosing between 'AI kills graduate jobs' and 'AI creates graduate jobs,' it is more accurate to watch how the expected productivity and proof at the point of entry are changing.

Candidates should ask what AI produces inside the company, not just whether the company uses AI

Even among technology employers, graduate demand can differ depending on whether AI mainly compresses internal work or is implemented for customers as a revenue-generating service.

For candidates, saying 'I can use AI' is becoming weak evidence. A stronger signal is showing what problem AI helped solve, what judgment remained human, and how the output was verified.

For IT services roles in particular, it can be useful to look beyond model knowledge and identify where experience connecting AI to existing systems, data, cloud and business workflows is actually required.

BANSEOG VIEW | Look at the company's position before assuming AI's hiring effect

Current public evidence does not prove that the total number of graduate hires across Korea's IT services industry is higher than last year. What it does show is that several major firms are recruiting graduates in the second half of 2026 while raising expectations around AI-use capability.

Placed next to the reconsideration of mass graduate hiring in parts of the platform sector, the evidence suggests that AI does not push every employer's entry-level hiring in the same direction.

A company using AI to compress existing work and a company paid to implement AI into clients' operations face different workforce equations. Instead of asking only whether AI destroys jobs, it may be more useful to ask whether AI is a cost lever, a productivity tool, or a new revenue-producing business for that employer.

Entry-level roles may not be disappearing uniformly. Their starting point may be moving toward people expected to solve real problems faster with AI.