The Gap with Samsung and SK Widens... SMEs Sit on a Decade of Data with No One to Use It [Everyone's AI, Divergent AX]①

Despite a Decade of Data, AI Talent Remains Scarce
Year One of AX: Small Businesses Are Left Behind

Editor's Note
Artificial Intelligence (AI) is transforming productivity and work processes for businesses, but there is a growing gap in the adoption speed between large corporations and small and medium-sized enterprises (SMEs). While companies equipped with dedicated AI teams and infrastructure are accelerating company-wide AI transformation (AX), those without such resources—including many organizations—are finding themselves left out, stuck in the AX blind spot and excluded from "AI for all." The Asia Business Daily examines the AX gap unfolding behind the scenes of innovation, and explores what it will take to achieve truly inclusive productivity gains with “AI for everyone,” leaving no one behind.

“We lack talent capable of leveraging the data we’ve accumulated over the past decade with AI. Despite the vast amount of data, our use of AI is still very basic.”

On the 7th, Lee Oseon, the CEO, is explaining the smart factory system and the use of AI at the Dong-A Plating office in Gangseo-gu, Busan. Photo by Eunseo Lee.

Dong-A Plating, a company specializing in automotive parts surface plating, participated in the 'Large and Small Enterprises Win-Win Smart Factory' project run by the Ministry of SMEs and Startups and Samsung Electronics in 2018, and is recognized as a business that has succeeded in digitally transforming (DX) its production site. By building a smart factory, Dong-A Plating improved productivity by 37% and achieved a dramatic result of reducing defects by 77%.

When the reporter visited Dong-A Plating's plating plant in Gangseo-gu, Busan, on the 7th, automated equipment was operating nonstop, continuing the process of treating the surfaces of bolts and nuts. Production line monitors displayed information like clients and processing times in real time. Instead of having to operate each machine directly as in the past, workers now set up operations via kiosk screens and check operating status as they move around the line.

Data collected in the plant is consolidated across four monitors in the administrative office. Information like production volume, defect records, and operation times for each plating line was being compiled in real time.

On the second floor, several employees were gathered to analyze root causes of product defects using AI tools. In another area, AI was downloading cost and annual production data from the system to automatically generate cost reports. Dong-A Plating is pursuing an AX project so that AI can optimize production conditions for each product automatically, but currently, their use of AI remains basic. CEO Lee Oseon lamented that the absence of a dedicated AI team makes it challenging to advance their smart factory with AX.

CEO Lee explained, “Existing management staff have to juggle their main duties with AI tasks, and even when trying to collaborate with external AI firms, it’s not easy to form a dedicated task force (TF) to work together. We urgently need specialists to teach field-specific know-how and experience to AI as domain expertise, and to research how to best apply it to each process.”

While agreeing that AX is necessary, Lee said that making a large-scale investment for AX is burdensome—especially at a time when survival itself is uncertain. “Even holding on for a day or a month is tough for SMEs, so there is less enthusiasm about AI than before,” he said. “AI may be a hot topic, but for SMEs, AX still feels a long way off.”

AI Advancement Requires Trial and Error...A Steep Path for SMEs

“It took over a year and a considerable amount of money to get here. It’s true that you need to go through trial and error to advance AI, but for SMEs, a single failure makes it hard to try again.”

Members of Wantedlab are conducting the in-house training program 'AX Champion,' where they directly solve work inefficiencies using AI. Wantedlab.

At Wantedlab, where 80% of employees develop and use AI tools themselves and the company has launched new businesses supporting AX based on its own AX experience, the journey has been anything but easy. Wantedlab operates ‘Wanted,’ an HR tech recruitment platform connecting job seekers and companies. Today, leveraging its successful internal AX, the company offers a packaged service helping corporate clients adopt AI.

In the early stages of AI adoption, the company keenly felt its staffing limitations. While requests to develop workplace AI tools increased throughout the company, the development team was too focused on adding AI features to the recruitment platform to keep up. Instead of simply hiring more developers, Wantedlab created an environment where frontline employees could directly improve their work using AI.

After declaring its AX initiative, it took a full year just to cultivate an AI-friendly environment. Streamlining the security certification and access rights system so company data—centered on resumes and job posts—could be used by AI took another six months. Existing personal learning stipends were converted into AI growth support funds, and 'token efficiency,' rather than simply 'token maximizing,' was adopted to control costs.

Jung Kisoo, Head of AI at Wantedlab, said, “Integrating AI tools into daily work requires significant time and money. You have to actually use AI and go through trial and error to discover how and where it’s effective, but for SMEs, keeping up investment in this process is difficult,” he said. He added, “AX is not lagging at SMEs because they don’t need AI; it’s a matter of lacking access to resources—cost, talent, data, education. Ultimately, it’s about enterprise-wide AI accessibility.”

Large Corporations Mobilize Entire Organizations...and the AX Gap Widens

In contrast to SMEs and organizations lacking data, talent, and organizational readiness—now increasingly relegated to the AX blind spot—large corporations, bolstered by ample investment capability, are fully mobilizing human and capital resources for company-wide AX.

Samsung declared a major AI transformation in June, rolling out generative AI services across all affiliates. The company is broadening the scope of AI adoption—from development to manufacturing—and is setting up dedicated teams for each affiliate to establish AX strategies and cultivate AI talent. SK has also convened its management to discuss AX implementation. At its subsidiary SK Telecom, employees are building their own agents to use for daily work and are able to develop workflow-specific AI through an in-house platform, with ongoing AI education provided throughout the year.

According to a report released by the Korea Chamber of Commerce and Industry last June, there is a 13.8 percentage point difference in the simple use rate of generative AI between large corporations (66.5%) and SMEs (52.7%). The proportion of SMEs reporting that they lack any generative AI roadmap was 70.4%, far higher than the 54.4% for large corporations. The gap was especially pronounced in manufacturing, with a difference of 24.2 percentage points, compared to 9.2% in the service industry—a ratio of 2.6 to 1.

The '2026 Corporate AX Benchmark Report' from edutech company TeamSparta also highlights the AX disparity between large firms and SMEs. When 330 decision-makers for AI training were surveyed on AX maturity, large IT companies scored the highest at 68, while manufacturing/production SMEs rated lowest at 28—a 40-point gap. While 37% of large businesses had advanced to the AI proliferation stage, over 50% of mid-sized and small businesses remained at the interest and consideration stages.

The investment gap was substantial as well. While 41% of large companies invested over 50 million won (KRW) per month in AI, only 9% of mid-sized companies did. The report noted, “Company size initially determines both starting point and investment capacity. This difference leads directly to a productivity gap and the later AX is adopted, the higher the catch-up cost becomes.”

This content was produced with the assistance of AI translation services.