South Korea Leads Global AI Safety Standards with Red Team Testing and Trustworthiness Labels

Korea Takes Leadership in Global AI Safety Standards

South Korea is positioning itself as a global leader in artificial intelligence safety and governance through groundbreaking international standardization efforts. The Electronics and Telecommunications Research Institute (ETRI) has proposed two critical standards to the International Organization for Standardization (ISO/IEC): AI Red Team Testing and Trustworthiness Fact Labels, marking Korea's transition from a technology "fast follower" to a "first mover" in setting global AI rules 1

1

AI Red Team Testing Standard Development

The AI Red Team Testing standard, designated as ISO/IEC 42119-7, represents a proactive approach to identifying vulnerabilities in AI systems. This methodology involves aggressively exploring and testing AI security by identifying scenarios where generative AI might produce incorrect information or be exploited to circumvent user protections 1

1

Source: Tech Xplore

ETRI serves as the editor for this international standard, developing common test procedures and methods applicable across diverse sectors including medicine, finance, and national defense. The institute has established partnerships with major Korean companies including STA, NAVER, Upstage, SelectStar, KT, and LG AI Research Institute to strengthen cooperation on international AI red team standardization 2

2

Medical AI Safety Initiative

In a significant milestone for Asian AI safety, ETRI collaborated with the Ministry of Food and Drug Safety to host Korea's first "Advanced AI Digital Medical Product Red Team Challenge and Technology Workshop" at the Novotel Seoul Dongdaemun Hotel on September 4-5. This event marked the first of its kind in Asia for advanced AI medical devices, bringing together medical professionals, security experts, and the general public to examine AI biases and risks 1

1

ETRI is developing medical-specific red team evaluation methodologies in collaboration with Seoul Asan Medical Center, creating comprehensive test systems for digital medical products that incorporate advanced AI technology and conducting empirical testing to ensure safety standards 2

2

Related Stories

South Korea launches world's first comprehensive AI law amid startup concerns over compliance

Korea Launches AI Safety Institute to Address Risks and Foster Global Collaboration

AI Leaders Call for Global Standards as US-China Competition Intensifies Over Development Controls

Global Tech Coalition Develops First International Standard for Large Language Models

Trustworthiness Fact Labels for Consumer Transparency

The second major standard, Trustworthiness Fact Labels (TFL), functions similarly to nutrition labels on food products, providing consumers with at-a-glance visualization of AI system trustworthiness. ETRI leads the development of the ISO/IEC 42117 series, which can be implemented through company self-reporting or third-party verification and certification 1

1

The TFL standard is designed to work in conjunction with the AI Management System Standard (ISO/IEC 42001), serving as a framework for organizations to demonstrate the trustworthiness of their AI-powered products and services. ETRI is considering incorporating Environmental, Social, and Governance (ESG) factors, including AI's carbon footprint, into future iterations of the standard 2

2

Strategic Alignment with National AI Policy

These standardization efforts align with Korea's Sovereign AI and AI G3 leapfrog strategies, representing a shift from merely securing technological capabilities to actively contributing to global AI governance leadership. The initiative parallels the role of the National Institute of Standards and Technology (NIST) in the United States, supporting national AI strategy through international standardization efforts 1

1

Kim Wook, Project Manager at the Institute of Information & Communications Technology Planning & Evaluation (IITP), emphasized that "providing AI safety and trustworthiness will make it easier for everyone to use AI, and leading the way in international standards this time is a turning point toward becoming a country that leads AI norms" 2

2

Lee Seung Yun, Assistant Vice President of ETRI's Standards & Open Source Research Division, noted that these standards address key technical elements in AI regulatory policies across the United States, European Union, and other nations, positioning them as common criteria for global AI safety and trustworthiness evaluation 1

1