Beyond experimentation
Siemens CEO Roland Busch and Naver CEO Choi Soo-yeon called on companies to move beyond artificial intelligence pilot projects and focus on measurable productivity gains. According to a report by the Korea Herald, the two executives spoke on Saturday at an industrial AI session at APK 2026 in Seoul, moderated by journalist Jennifer Moon: the ability to unlock internal data and integrate AI into daily operations, they said, will determine the technology's economic value.
The stance marks a shift in emphasis from the experimental phase: the criterion indicated is not adoption itself, but verification of results. The statements, as reported by the source, do not amount to verified data on companies' actual performance.
Data, leadership and measurement
Choi argued that companies must first organize information scattered across paper archives, employee computers and departmental systems into formats usable by AI, ensuring its accuracy, currency and access for authorized users. She also warned that producing reports or other materials with AI does not automatically translate into greater productivity: companies should measure whether the technology reduces waiting times, improves efficiency or eliminates repetitive tasks.
Busch emphasized the role of management and ambitious goals to encourage adoption across the company, citing the case of a Japanese automaker that, with Siemens technology on 10,000 machines, reportedly achieved a return on investment in less than three months. This is an example provided by the CEO, not independently verified data.
Choi cited Naver's experience with more than 200 delivery robots active for over four years at its 1784 headquarters, describing it as a demonstration of the difficulties of delivering reliable AI-based services in environments shared with people.
Interoperability, sovereignty and labor
Data sharing remains a significant obstacle: in an audience poll, 41% of respondents identified data sharing and interoperability as the main barrier to the spread of industrial AI across regions. Busch called for clear policies on the matter and specialized skills, noting that customer data could be used to train industrial AI models without necessarily leaving the company's premises.
On technological sovereignty, Choi said countries and companies can collaborate on semiconductors, AI models and services while maintaining control over their data. Busch pointed to possible synergies between Siemens' chip design technologies and Korean semiconductor manufacturers, as well as opportunities in life sciences and drug development.
On employment, Busch warned that repetitive and rules-based jobs will be exposed to replacement, calling for reforms in education and training. Choi said Naver continues to hire every year but will need new business models to offset the growing costs associated with AI. Busch later visited Naver's headquarters, confirming the partnership between the two companies, which includes Siemens fire protection systems and involvement in the infrastructure of Naver's Gak data center.
What remains to be verified
The material reports statements and an audience poll, without independently verified performance data or details on the time and investment required for smaller companies. It therefore remains unclear how the proposed measurement criteria translate into practice and which data-sharing policies are actually adopted.
