A worker moves a finished insulating glass unit with a vacuum lifter at a glass plant in Cheongju, North Chungcheong.KIM KYUNG-MI
Large companies in Korea are adopting AI at more than twice the rate of smaller businesses. Manufacturers and industry groups say the gap could become an "AI divide" running through the country's manufacturing supply chains.
A total of 77.3 percent of large companies surveyed were already using AI at work, against 28.9 percent of small- and medium-sized enterprises, according to a JoongAng Ilbo survey conducted by Hankook Public Opinion Research in late May. The survey covered 327 large, mid-sized, small and state-run companies nationwide, with a margin of error of 5.42 percentage points at a 95 percent confidence level. On a factory floor, AI transformation (AX) means buying equipment, installing software and hiring specialists.
One owner who has run a window and door manufacturing company for 10 years considered installing an AI vision system on the factory floor early this year. The system would have read the dimensions off a blueprint, cut the glass to size automatically and reported the result. The quote came back in the hundreds of millions of won, and the owner dropped the plan.
"For manufacturers, one piece of AI equipment costs 50 million won [$36,200] and installing the AI software costs 80 million won," Kim Gwang-hu, an executive vice president at the AI consultancy DX Solutions, said Thursday. "AX is not possible without a baseline investment of 300 million won, so it is a heavy burden for a small company. Dedicated AI staff also cost money, so the transition is bound to be slow."
TeamSparta, which sells AI training to companies, found that 38 percent of large companies surveyed had budgeted at least 50 million won for AI training, according to a benchmark report published Aug. 6. The figure compared to 14 percent of mid-sized companies and less than 5 percent of small ones. The report drew on responses from 330 people who decided on AI training at their workplaces.
"Since the beginning of the year, we have been told to use AI actively, but the company has never provided training or explained specifically how to use it," an employee at a small manufacturer in Seoul said. "At first, it paid for subscriptions to services such as ChatGPT, but even that stopped as the economy worsened."
A worker fits an aluminum spacer frame between panes on an insulating glass line at a plant in Cheongju, North Chungcheong.KIM KYUNG-MI
Large manufacturers already send suppliers the production data and quality standards that their own AI systems have processed. The suppliers have neither the systems nor the staff to handle it, which leaves a bottleneck at the bottom of the supply chain, where good data cannot be put to use.
"Many companies have not even completed basic digital transformation, so they have nowhere near enough data, and in many cases the data they do hold has never been classified," Lee Ji-yeon, head of AI innovation at the Korea Federation of Small and Medium Business, said.
The Federation of Korean Industries released a survey on Aug. 13 covering Korea's 800 largest manufacturers by sales, to which 200 companies responded. Of those, 91.5 percent said that AX across the entire supply chain, not just at individual companies, mattered to manufacturing competitiveness. When asked what was blocking them from supporting suppliers, 43.4 percent pointed to suppliers' weak digital infrastructure and limited AI skills, the most common answer, and 17.5 percent to the cost of adoption.
"The central government, local governments, universities and other expert groups, as well as large companies, need to work together and build the foundation for AX across entire supply chains," Choi Byung-ho, a professor at Korea University's Human-Inspired AI Research Institute, said.
This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom.
