As Korean employers bring AI into recruitment, some are testing whether applicants can check its work, not merely use it. One example is Nexon dropping a coding test for an internship programme in favour of AI-based interviews and practical assignments; the broader shift raises a useful question about what entry-level candidates are now expected to know.
Watch Desk analysis
What happened
The Korea JoongAng Daily describes several different approaches. MyRealTrip is recruiting entry-level staff without assigning them to particular roles or departments in advance, and says it has no university-major restrictions. Its selection process asks applicants to show how they have solved problems, then assesses their reasoning, decision-making and use of AI tools.
Nexon uses AI-based interviews to assess applicants’ programming foundations, followed by practical assignments that examine how they use AI. The report also says Doosan asks applicants to produce a report with AI in 30 minutes, while Kia assesses how candidates use AI to analyse and solve problems. These are company-specific examples, not a single standard adopted across Korean employers.
Why it matters
The useful distinction is between getting an AI tool to produce an answer and knowing whether that answer deserves to be trusted. Employers in the report are looking for both tool use and independent judgement, a combination that may be harder to demonstrate when applicants have little workplace experience.
It also changes what a hiring test can reveal. A polished AI-generated answer may say less about a candidate than how they frame the task, check the result and explain their decisions. The report notes that employers are taking different routes, so there is no settled playbook yet.
Our read
This is a more interesting hiring signal than another demand that applicants “know AI”. Testing how people verify output gets closer to the skill employers say they need. But there is a fine line between assessing judgement and quietly expecting graduates to arrive with experience they have not had a chance to gain.
For applicants, the practical message is to be ready to explain how they used AI and how they checked its work, not just to show off a clever prompt. For employers, a test is only useful if it measures the job rather than who has the most practice with the test itself.
What to watch
- Whether more employers add AI-assisted tasks to recruitment.
- How companies distinguish independent reasoning from a polished AI-generated response.
- Whether universities adapt courses to teach practical AI use alongside verification.
- Whether these tests widen the gap between employers’ expectations and graduates’ experience.
Discussion spark: Should entry-level applicants be expected to prove they can verify AI output before they have workplace experience, or should employers teach that skill on the job?
Sources and evidence
- Source update (26 October 2026, 22:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.