국내 AI 전문가들 “개발 제동보다, 안전성 검증 속도 높여야”

South Korean AI specialists argue that it is more important to accelerate development speed and tighten safety validation rates than to slow down development progress. At a seminar in Seoul, the National Artificial Intelligence Strategy Committee discussed the risks and response capabilities of frontier AI models. The committee's chair emphasized the need to maintain development speed while rapidly improving control, recovery, and accountability.

He noted that speeding up development does not mean stopping development; rather, it means securing time for control. The chair called for Korean participation in creating international safety standards, warning that a slow-down in development could deepen technological dependency. The chair also highlighted the need to fill technological and institutional gaps in AI evaluation and control, such as verification, combatting misuse, containment, and tracking.

The importance of developing independent evaluation standards and risk judgment capacities in Korea was also emphasized. Concerns about a potential "regulatory dominance" by US big tech were raised, along with the need for international cooperation through organizations such as the UN and the OECD in establishing AI safety standards.