A survey has revealed that while Korea is considered a leader in the Asian region regarding the adoption of artificial intelligence (AI), challenges remain in the expansion phase concerning workforce, operational structure, and regulatory compliance. In particular, although the achievements of AI projects have been confirmed, a gap was found in the process of disseminating them throughout the organization.

ST Telemedia Global Data Centres (STT GDC) announced on the 27th that it has released the results of a survey on Korea as part of its report, 'Mind the Gap: Bridging the AI Infrastructure Readiness Divide,' which analyzes the status of AI infrastructure readiness in Asia. The survey was conducted by Ecosystm and involved over 600 corporate and digital organization leaders across nine Asian countries, including Korea.

According to the report, Korea was analyzed to be among the top in Asia in terms of AI adoption levels. However, based on infrastructure maturity, the early and intermediate stages accounted for a high proportion.

The breakdown was 1% for the exploration stage, 67% for the implementation stage, and 32% for the integration and leading stage; in particular, the proportion of companies in the leading stage remained at approximately 2%.

This can be interpreted to mean that while they are proactive in adopting AI, the extent to which they translate it into overall organizational competitiveness is limited.

On the other hand, 75% of respondents answered that AI projects are generating value beyond expectations. This figure significantly exceeds the Asian average of 34%. The report analyzed that this indicates domestic companies have already confirmed the results of AI utilization.

Difficulties in the AI expansion process were also identified.

52% of respondents stated that they lacked the internal expertise to manage and optimize high-density AI infrastructure, while 48% identified the burden of initial investment and operating costs as a major challenge. 52% cited data sovereignty and regulatory compliance as constraints.

The explanation is that operational complexity and the issue of securing specialized personnel are acting as major obstacles. It is analyzed that a capability gap has emerged during the actual operation phase of AI infrastructure.

A gap between perception and practice was also observed regarding sustainability. While 31% of responding companies stated that they actively consider sustainability when making decisions about AI infrastructure, the weight given to this factor in actual colocation provider evaluations was only about 15%. Similarly, although approximately 48% of respondents said they were reviewing or adopting liquid cooling technology, the survey found that it was not sufficiently reflected in partner selection criteria.

The report analyzed that changes are also appearing in the methods of securing infrastructure. Companies moving from the integration stage to the leadership stage tend to choose to secure expertise through strategic partnerships rather than possessing their own infrastructure.

Heo Cheol-hoe, Representative of STT GDC Korea, stated, “Korea has moved beyond the AI experimentation stage,” adding, “In the future, the key challenge will be how to expand and operate the established AI environment.” He further added, “Partnerships to secure expertise and infrastructure design that considers the regulatory environment will become crucial.”