SGA솔루션즈 CEO proposes AI controls within existing Korean security certifications
At a 한국정보보호학회 workshop on October 22, SGA솔루션즈's CEO urged adding AI-specific controls to ISMS-P, N2SF and K-RMF rather than creating a new certification.
SGA솔루션즈 CEO 최영철 proposed a Korean AI cybersecurity framework that adds AI-specific controls to existing national security systems and cross-links duplicate standards across ministries, in a talk at the 한국정보보호학회 '2026 AI x RMF 워크숍' held on October 22 at Samsung SDS West Campus in Seoul, according to dailysecu.com.
최영철 argued against creating a separate new AI security certification. Instead, he said AI-specific controls should be folded into the systems Korea already runs, noting that the country's security regime is split across the public, private, defense and financial sectors, each with different controls and procedures.
As a model, he cited the U.S. NIST CSF and RMF, which sit on the shared control baseline SP 800-53. He structured his proposed framework around three areas — governance, threats and controls — and named prompt injection, training data poisoning, AI model compromise and AI agent privilege abuse as the AI threat types it would need to address.
On implementation, he suggested adding AI controls to ISMS-P, N2SF and K-RMF, together with common control ID numbers and cross-mapping tables so a single measure can be recognized across multiple certification schemes.
최영철 stressed that the presentation was his personal proposal and does not represent confirmed government policy.
Quick answers
Who proposed the Korean AI cybersecurity framework?
SGA솔루션즈 CEO 최영철, speaking at the 한국정보보호학회 '2026 AI x RMF 워크숍' in Seoul on October 22.
Which Korean certification systems would the proposal change?
He suggested adding AI controls to ISMS-P, N2SF and K-RMF, with common control IDs and cross-mapping tables.
Is this confirmed government policy?
No. 최영철 said the proposal was his personal view, not confirmed government policy.