en.Wedoany.com Reported - South Korea's revised enforcement decree of the "Basic Act on the Development of AI and Building Trust, etc." took effect on the 21st, requiring public institutions to prioritize the procurement of artificial intelligence (AI) products and services. This move is expected to create greater opportunities for domestic AI companies to enter the public market.

Under the revised decree, the Ministry of Science and ICT will review AI products and services actually used by public institutions and issue confirmation certificates. Products that receive the certificate can enjoy various policy benefits, including relaxed requirements for participating in Multiple Award Schedule (MAS) contracts, additional points in total price contract qualification reviews, and exemption from supply performance requirements for software (SW) unit price contracts. The government aims to stimulate diffusion in the private market by making public institutions initial demanders of AI products. The industry has welcomed this, viewing it as opening new sales channels for emerging AI companies lacking public procurement experience.

However, some remain cautious. Critics point out that if public institutions fail to clearly define the specific problems to be solved and performance goals beforehand, and hastily introduce AI products solely based on the "prioritize AI" policy direction, projects may become superficial, leading to a "face-saving AI" phenomenon. The industry calls for ensuring that obtaining the confirmation certificate translates into actual procurement and use, preventing the system from becoming just another certification process for companies.

The industry generally believes that the AI product and service confirmation system could change the current contracting model for public AI projects. Previously, public AI projects often involved developing systems tailored to each institution's needs. In the future, procurement centered on technically verified AI products and platforms may increase. A representative from an IT service company stated that for domestic AI companies with existing procurement registrations and public reference cases, market entry opportunities will significantly expand. Emerging companies will also benefit from some relaxed requirements, using the confirmation certificate as proof of technical reliability. The industry also expects that supply cases accumulated in the public sector can lay the foundation for companies to enter private markets such as finance, manufacturing, and distribution.

However, as the system is just starting, most companies remain cautious. Even if the number of companies obtaining confirmation certificates increases, doubts remain about the actual impact on sales growth if public institutions fail to secure relevant budgets and actually issue projects. The connection between the confirmation system and existing procurement systems is also crucial. If it is not linked to actual procurement, companies may view it as just another procedure. Experts suggest operating product confirmation standards objectively and transparently, not only verifying whether products have AI functions but also assessing their performance, stability, and continuous operation capabilities required in actual business contexts.

Experts further point out that rather than prioritizing AI product review, public institutions should first clarify what problems they need to solve. The public service sector emphasizes stable operation over pioneering new technologies, as failures in areas like administration, welfare, and security directly impact citizens. Professor Kim Sook-kyung from the School of Technology Management at the Korea Advanced Institute of Science and Technology (KAIST) emphasized that the public sector must provide stable services, so it cannot solely pursue technological leadership. Instead, it should first define task objectives and then integrate the most suitable and stable technology. She expressed concern that the "prioritize AI" signal might be interpreted as incorporating AI into more projects, potentially leading to technology introduction in areas where it is unnecessary or advancing projects without adequate institutional preparation.

Numerous cases show problems with the design of Key Performance Indicators (KPIs) for measuring project performance. Some projects only present technical figures like accuracy rates or automation ratios, failing to verify whether AI has genuinely improved administrative efficiency and service quality. A policy expert noted that clear objectives are necessary to select appropriate technologies and ensure continuous operation. If the focus is solely on introducing AI, the introduction itself becomes the goal, while the reason for introduction is overlooked. Therefore, before promoting digital transformation (AX) in the public sector, existing business processes should be redesigned, clearly distinguishing tasks performed by AI from those requiring human judgment, and predetermining the authority, responsibility, and control procedures between the two. With the proliferation of AI agents, issues of responsibility arising from authority misuse, bias, and erroneous judgments will become more prominent. Experts analyze that the success of the AI Basic Act ultimately depends on whether it produces tangible results in the field, not on the number of confirmation certificates issued or projects launched. The core question is not "whether public institutions purchased AI," but "why AI is needed," "how existing tasks will change," and "who will operate and be responsible after technology introduction."