AI 역량 테스트를 바라보는 두 가지 관점 - 뉴스H

AI competency test enables to identify applicant propensities in a short period of time There are also negative opinions due to lack of reliability in AI competency information and evaluation criteria The AI competency test, which was introduced in Korea in 2017, has emerged rapidly as contactless recruitment methods have increased since COVID-19 pandemic. According to Midas HRi, an AI competency testing company, the number of companies that introduced AI competency tests in 2021 increased by 50% compared to the previous year to 450. Currently, large companies and public institutions are conducting AI competency tests for hiring employees. Hanyang University has also introduced AI competency tests in recruitment since July 2020. "I felt the need for contactless interviews along with simplified recruitment steps in 2020 because awareness of COVID-19 was at its peak," said Lee Jong-hyun, an employee of the Human Resource Team, adding, "In this context, we changed the second stage of screening from aptitute test to AI competency test." How does AI competency test work? The AI competency test typically consists of 'identifying propensity', 'video interview', and 'strategic game'. The propensity identification evaluates morality and accountability, analyzing answers to multiple-choice questions. In video interviews, applicants answering the subjective questions are recorded to analyze voice, expression, and speed of answers of interviewees are analyzed. Through this process, the ability to express opinions, confidence, and reliability are evaluated. Finally, the strategic game judges the candidate's cognitive and analytical skills by letting them perform games such as 'rotating shapes'. Scores and rankings that comprehensively evaluate these items are delivered to human resources managers and used in the recruitment process. Range of AI competency testing use Although there are differences between companies, AI competency tests are still being used to refer to applicants' propensities. Lee said, "Even when hiring Hanyang University employees, AI competency tests do not account for a large portion of the entire process." He added, "Even if they accounts for small portion, there is a process of verifying the test results in in-person interviews. It is better to respond as conscientiously as possible." On the other hand, some companies place great importance on eliminating applicants from the hiring stage if the AI competency test does not reach a certain score. Thus, applicants should carefully check the detailed scores. Advantages of AI competency testing The AI competency reduces not only in-person contact between recruiters and applicants, but also problems related to fairness in recruitment. In particular, its strength is notable in that it can efficiently select applicants as the number of ad-hoc hiring increases from time to time. This is because it helps the human resources managers to automate the work of the evaluating applicants in the era of the universalization of occasional recruitment. "It is useful in that it can identify the capabilities of many applicants in a short time and can be used as a reliable reference for in-person interviews," Lee said, adding, "I think the advantage is that it can identify applicants based on AI big data analysis without prejudice from education and specifications." For applicants, the positive aspect was that they could have an interview in a comfortable place. This is because they feel less nervous as there is no place restrictions. In addition, the fact that it is more reliable through data-based evaluation, not a person who has subjective standards, also acts as a factor that positively views AI competency tests. Problems of AI competency testing and future improvements Contrary to positive opinions, there are voices of concern that AI competency tests will give burden on job seekers as it gets more common. Since the AI competency test has been conducted in the public for a relatively short period, the cost burden and anxiety are added to job seekers, since they have to attend private academies due to a lack of information. In addition, there are many opinions that it is difficult to believe the results of the AI competency test. Due to the absence of sudden questions or unexpected questions, it is possible to take the test in a staged manner that is easier to pass the AI competency test rather than the applicant's real self. Accordingly, plans are needed to reduce the difference in views between companies and applicants. In particular, it is necessary to present accurate standards such as evaluation methods and data utilization methods so that applicants can trust the interview evaluation method. In addition, companies should secure and analyze high-quality data through algorithm management to provide trust to applicants. If companies make various efforts to conduct objective and fair recruitment, they will be able to reduce applicants' negative views on AI compete