[Edaily Reporter An Yu-ri ] Hiring practices for developers in the information and communications technology (ICT) industry are changing. An increasing number of companies are eliminating “coding tests”—once considered the standard entry requirement for developer recruitment—and instead assessing candidates’ ability to work alongside artificial intelligence (AI).

Nexon has decided not to administer coding tests for “Nexorial,” its recruitment-oriented internship program for this year’s new graduate developers, for which applications opened on the 25th. Instead, Nexon has introduced a new “AI Utilization Competency Assessment” phase.

This process comprehensively evaluates candidates’ understanding of AI, problem-solving approaches, and structural thinking as they use AI tools to solve problems similar to real-world tasks. Nexon stated that it has significantly overhauled its hiring process as the ability to utilize AI in practical work environments has become increasingly important.

Changes are also emerging in the recruitment of experienced developers. CJ Corp. introduced an “AI assignment” track in place of the traditional live coding test during its open recruitment for experienced developers in the development division last July. For the AI assignment, the company utilized “Cofa-Probe,” a recruitment and evaluation solution developed by KRAFTON(259960).

CJ Corp. and KRAFTON also held a joint AI hackathon last month. Under the slogan, “Don’t just take a coding test—prove how you work with AI,” the hackathon assessed developers’ capabilities by having them plan and implement a service within a limited timeframe using AI tools.

A KRAFTON spokesperson stated, “KRAFTON is currently applying Cofa-Probe to the recruitment process, primarily for development roles, and plans to expand its scope in the future.”

A coding test is a programming exam that evaluates a candidate’s understanding of algorithms and coding skills. It has long been regarded as the first hurdle in the initial stage of developer recruitment at major companies. Conducted both online and offline, it has been widely used, particularly in large-scale public recruitment drives.

The recent decline in coding tests can be attributed to the proliferation of generative AI. In the development field, using AI to write code has long been commonplace. Actual productivity is now determined not by how quickly one can write code manually, but by the ability to instruct AI appropriately and judge whether the results are accurate.

This shift has also been influenced by a general slowdown in hiring; unlike the 2020–2022 period, when coding tests were widely adopted, large-scale public developer recruitment drives have become rare.

An official from a mid-sized software company stated, “There is a widespread industry perception that coding test scores are not strongly correlated with on-the-job performance,” adding, “The ability to understand structure and solve problems is more important than the ability to write code itself.”

In its report titled “Prospects for Changes in Software Developer Hiring and Generative AI,” the Software Policy Research Institute noted, “While demand for entry-level developers is decreasing, demand for senior developers capable of utilizing AI technology is on the rise,” adding, “There is a growing need for entry-level developers to redefine their roles by collaborating with AI.”