On the 27th, at the National Assembly Members’ Office Building, participants posed for a commemorative photo at a seminar titled ‘Era of AI Transition, Status of the Software Industry and Survival Strategies for Software Companies.’ [Digital Daily, Reporter Lee Anna] As the transition to artificial intelligence (AI) accelerates, the software industry is experiencing varying degrees of crisis. While some companies have turned productivity improvements into opportunities, others face reduced employment and market encroachment. The 30-year-old public IT contract structure is blatantly revealing its limitations in the AI era.
At the ‘Era of AI Transition, Status of the Software Industry and Corporate Survival Strategy Seminar’ held by the Software Association under the auspices of Representative Kim Hyun of the Democratic Party on the 27th at the National Assembly Members’ Office Building, software company representatives and experts from academia and the legal field gathered to voice their concerns.
The most distinct shared concern among attendees was employment issues. As AI quickly absorbs tasks from code writing to testing and documentation, the recruitment of new developers has notably decreased. Bae Hyun-seop, CEO of SureSoftTech, mentioned that while they have been increasing their workforce by over 50 people annually, last year they only hired 30. Although they continue mandatory public recruiting, considering AI productivity, they suggest that it might be feasible to reduce hiring even further.
Yoon Gi-hyun, CEO of UNOPoint, also mentioned that they have barely recruited new employees over the recent two years. Shin Jung-kyu, CEO of Laybelumb, explained, “This winter, the interns left without writing a single line of code by hand,” indicating that the role of developers is shifting from coding itself to designing the process that allows AI to handle coding. The adoption of AI is directly leading to reduced job opportunities.
However, there was disagreement regarding the impact of AI on the existing software market. Song Ho-cheol, head of the Doosan BizOn Platform Business Division, stated, “We recorded the highest sales ever since launching the AI agent service.” He mentioned that AI agents connecting and executing services on existing ERP, HR, and accounting systems have increased the utilization of existing software.
In contrast, MarkAny’s CEO described meeting with representatives from ten surrounding startups, all of whom were removing their Software as a Service (SaaS) offerings. MarkAny also replaced their project management tools by directly creating them with AI coding.
The differing viewpoints stem from the different markets each company belongs to. For platform-type companies that have accumulated domain knowledge and data for decades, AI is a tool to enhance existing services. In contrast, for other companies, it manifests as removing existing SaaS. The same technological change affects companies in entirely different ways based on their position.
Another issue raised was how to distribute the benefits from productivity enhancement. SureSoft saw nearly a 60% rise in operating profit after adopting AI last year. CEO Bae is restructuring contracts to transparently share these gains with clients, advocating that simply reducing prices would remove the incentive for software companies to invest in productivity improvements.
The seminar also addressed the public IT market’s contract structure. In public software projects, budgets are calculated using the Function Point (FP) system, but contracts are based on Request for Proposals (RFPs). Since these two criteria do not match precisely, discrepancies in requirements interpretation during the build process can lead to project scope disputes. Clients want better systems and increase requirements, while contractors must fulfill the project within the original budget, causing conflicts.
Although this issue is not new, the increase in disputes has reasons. With project scales growing from hundreds of millions to billions, the impact of disputes has widened, and companies, free from the pressure to win subsequent projects due to large enterprise participation restrictions, increasingly choose legal actions. Lee Jung-taek, vice president of ITCENENTech, proposed that formally including a clause prohibiting FP quantity excess in contracts would provide a basis for clients and contractors to negotiate within the same limits.
Chae Hyo-geun, vice chairman of the IT Service Industry Association, critiqued the fundamental structural issue, claiming, “RFPs cannot be clear. They are subjects of management, not engineering.” Although the Software Promotion Act already provides a basis for fair compensation, it is not translating into practice because different ministries, including the Ministry of Economy and Finance and the Ministry of the Interior and Safety, are not aligned with this law.
For small and medium-sized enterprises (SMEs), there are more realistic barriers than AI strategy formulation. Park Chang-won, senior executive at INNOGrid, mentioned, “Some places have a one-week validity period for GPU quotations with final prices determined at delivery,” depicting a scenario where even conducting AI experiments is challenging due to fluctuating costs. He requested that the government create an environment where SMEs can trial AI with some room for trial and error.
Representative Kim Hyun (Democratic Party) concluded, “In the AI era, the software industry can either face a crisis or an opportunity for resurgence.” He emphasized that if rigid contract practices are tackled and the value of technology is fully recognized, the AI transition will certainly be an opportunity. He added, “The role of the National Assembly is to connect the voices from the field to legislative and policy improvements,” pledging to develop the opinions raised that day into specific legal and institutional improvement tasks.