[AI Era, Jobs Are Changing] How AI Is Transforming the Labor Market... Are Institutions Ready?

AI Penetrates the Job Market First
AI Algorithms May Lead to Biased Hiring
US and Europe Advance Regulatory Frameworks... Korea's Basic AI Act
Remedies and Specificity Still Lacking

Recently, an unprecedented class action lawsuit has been filed in the United States against an artificial intelligence (AI) development company. The lawsuit, brought by job applicants against Workday, a developer of AI-based human resources management software, alleges that the company's AI algorithms systematically excluded older applicants, claiming age discrimination.

In May, the U.S. District Court for the Northern District of California granted preliminary approval for the case. The court's decision means that the plaintiffs' claims are subject to legal judgment, and as the main proceedings begin, the lawsuit is likely to enter a prolonged phase.

This lawsuit is significant because it goes beyond the controversy surrounding a single company, serving as the first case to institutionally determine accountability in the labor market of the AI era. There are growing calls for the establishment of legal standards regarding who should be held responsible, and to what extent, when AI is involved in all stages of labor, including recruitment.

AI is already deeply embedded in the recruitment process, which is the first gateway to the labor market. In South Korea, major conglomerates such as Samsung, LG, and Hyundai Motor Group, as well as financial institutions, are actively utilizing AI for document screening, online assessments, and AI interviews. While domestic companies emphasize that AI is merely a tool to assist with complex hiring procedures and that final decisions are made by humans, in reality, algorithms have a significant impact on rejections during the document screening stage.

Current labor and anti-discrimination laws do not adequately reflect these changes. The existing legal framework is based on the premise that humans evaluate other humans, but today, it is common for the decision-maker in the hiring process to be unclear. In some cases, applicants filtered out by algorithms are not even reviewed by human recruiters.

An even greater issue is that most AI hiring systems operate as black boxes. The criteria used by the algorithms are either protected as trade secrets or are technically difficult to explain. As a result, job seekers can only claim discrimination based on statistical outcomes, not on the grounds of individual decisions.

Consequently, there are increasing calls for institutional measures to ensure procedural transparency and remedies for those who have suffered disadvantages. At the same time, there is a need for social consensus on who should be held accountable when AI errors or biases occur. Key issues include whether the responsibility lies with the company that commissioned the recruitment, the software developer that designed the algorithm, or if it should be considered a simple technical error.

The United States has recognized these issues and is working to update its regulatory framework. Although there is no comprehensive federal AI hiring law, the Equal Employment Opportunity Commission (EEOC) interprets AI-based hiring tools as part of the "selection process" under the Civil Rights Act, the Americans with Disabilities Act, and the Age Discrimination in Employment Act. In certain regions, such as New York City, regulations require bias audits, disclosure, and advance notification to applicants for automated employment decision tools.

The European Union (EU) has gone a step further by introducing the AI Act, which regulates AI according to risk levels. AI used in recruitment, employment, and human resources management is classified as "high-risk AI," and is subject to mandatory pre-risk assessments, transparency requirements, and human oversight. In the event of an AI error, the user company bears primary responsibility, and if a structural defect in the algorithm is found, the developer is jointly liable. Job applicants who have suffered harm can seek correction and remedy under the "AI Liability Directive," which is currently under discussion, and the burden of proof rests with the company or developer.

South Korea is also at the starting line of this discussion. Alongside debates on amending the Fair Hiring Procedure Act, the "Basic Act on Artificial Intelligence," which regulates AI in general, passed the National Assembly at the end of 2024 and is set to take effect in January this year. However, critics point out that the law currently remains at the level of declarative principles and guidelines, lacking specificity regarding actual use in recruitment and remedies for discrimination in practice.

Hong Seongmin, a research fellow at the Korea Legislation Research Institute, said, "While closely examining the adoption of AI technology under the basic framework of labor law, it is necessary to establish separate legal systems for areas where collective interest adjustment does not function or where special protection is needed."