MSIT Cuts Tens of Billions of Won
Concerns Over 'Maltreatment' via Investment/Support Cuts
“Must Build AI Physical Fitness Through SW”
As the government announced another massive budget injection into artificial intelligence (AI) for next year, the budget for traditional software (SW), which serves as the foundation of the AI industry, was found to have taken a step back. While investment in foundational SW must go hand in hand as AI investment increases, voices of concern are growing that reduced support could worsen not only SW but also AI competitiveness.
According to the industry and the Ministry of Science and ICT budget plan on Sept. 8, among the non-R&D budget plan of the MSIT for next year, the SW-related budget was confirmed to be at the level of KRW290 billion, down by about KRW40 billion from this year's approximately KRW330 billion.
Looking into details, projects directly linked to SW industry competitiveness were reduced one after another. The budget for Expanding SW Industry Infrastructure (SW Enterprise Growth Facilitation Support, SW Industry Technology Expansion Capability Strengthening, etc.) was about KRW17 billion, down 15% compared to this year's main budget. During the same period, the XaaS Leading Project and Public SW Business Modernization project decreased by 28.5% and 15%, respectively. In addition, the SW Industry Overseas Expansion Capability Strengthening project was also reduced by about KRW500 million to about KRW3 billion.
The same is true from the perspective of the SW Bureau of the MSIT, which is in charge of SW projects. The budget for AI-Centric Universities, managed by the bureau, increased by about KRW67 billion from KRW117.96 billion this year to KRW185 billion next year, but excluding this, the SW project budget effectively decreased. New projects directly connected to the SW industry, such as SW system improvement and SW competitiveness enhancement, were also found to be almost non-existent.
Criticism is rising that as the government concentrates its budget on AI models and AI infrastructure, support for SW—which is the foundation of AI—has been pushed down the priority list. Industry experts argue that the shrinking role of the SW industry and developers due to AI coding, raised by some quarters, is a superficial view.
Chae Hyo-keun, vice chairman of the Information Technology Services Industry Association, worried, “The claim that SW becomes unnecessary when AI agents and others become active stems from a misunderstanding that views SW as simple coding.” He added, “In reality, AI cannot be implemented without database (DB) and middleware SW, and the physical AI that the government proclaims to leap forward as world No. 1 is also impossible without embedded SW.” He continued, “Shouting for an AI powerhouse and executing massive investments only in AI infrastructure without investing in these basic SW areas is like building a castle on sand.”
Voices are rising that it is time to focus more on competitiveness in the SW field, which gauges basic AI stamina.
To apply AI models and services to actual industrial and public fields, capabilities in cloud, DB, and development tools, as well as connection and operation with existing information systems, must be backed up, and all of these are parts implemented by traditional SW companies. Pointing out follows that budgets supporting cooperation between AI technology companies and existing SW/IT service companies, long-term operation and maintenance, and commercialization must be accompanied without fail.
Kim Sook-kyung, professor at KAIST, pointed out, “AX is not achieved just by adopting AI.” She noted, “Cloud conversion, data refinement, modernization of aging systems, and connection with existing systems must be carried out together, but budgets and policies to support this are not receiving sufficient attention.”
Prof. Kim emphasized, “AI must also ultimately survive in SW competitiveness and the SW market ecosystem, but the government should not shrink the existing market and only expand proof of concept (PoC) or R&D projects.” She added, “Because it is difficult under such a structure for AI to reach a level where it can properly commercialize and connect with existing systems, the government must provide maximum support for SW-related budgets, projects, and policies as well.”
