KT Corp leads AI cybersecurity standardization at UN body, aims to set 'digital trust' benchmarks

KT Corp has stepped up efforts to shape international AI cybersecurity standards at the International Telecommunication Union, a UN agency. The company said it plans to extend its standardization work into the digital trust arena to strengthen its leadership in information security in the AI era.

KT said Monday it led discussions on next-generation information security standards at ITU-T Study Group 17, the ITU's working group for developing international standards in information security. The group brings together information security experts from around the world to develop international standards covering cybersecurity, security architecture, personal data protection and AI security. The meeting was held in Chongqing, China, from Sept. 7 to Sept. 11.

Kim Chang-o, head of KT Corp's personal data protection group, attended as chair of ITU-T SG17 Working Party 3, which focuses on assurance and security strategy and cooperation. Yeum Heung-yeol, chairman of KT Corp's personal data protection advisory committee and a former ITU-T SG17 chair from 2016 to 2024, also participated in the effort.

According to KT, the growing spread of generative AI has drawn increasing attention to the importance of AI-based threat detection, analysis and response capabilities in cybersecurity. At the meeting, the company proposed guidelines for developing cybersecurity-specific AI models using general-purpose foundation models as a new international standards work item.

The contribution paper submitted by Kim and Yeum covered four areas: cybersecurity knowledge sources and datasets for AI model development, technical considerations for developing cybersecurity-specific models, evaluation considerations across the full model development lifecycle, and recommended practices for continuous model improvement.

KT said that establishing guidelines for developing cybersecurity-specific AI models would enhance the reliability of AI security solutions. The company added that such standards would give businesses a consistent framework for developing AI models while also improving the interoperability of AI security technologies globally.

KT said digital trust has also emerged as a key priority. ITU-T SG17 is currently developing a technical report that defines "trust" and "trustworthiness" in information and communications technology environments and systematizes the components of a trust framework for applying those concepts to systems.

KT has been contributing to the development of that technical report, proposing revisions that supplement key term definitions and add requirements to the trust framework. The report has been finalized with support from the United Kingdom and India.

Going forward, KT plans to expand its international standardization activities to encompass AI cybersecurity, personal data protection and digital trust. The company will also deepen cooperation with domestic and international experts and standards bodies, with the aim of ensuring that South Korean technology is reflected in global information security standards.

"Now that AI technology is spreading across all industries, it is important for the international community to jointly establish development standards for cybersecurity-specific AI models and trust frameworks for ICT services," Yeum said. "We will work to ensure that South Korea's experience is reflected in international personal data protection standards."

"In the AI era, cybersecurity, personal data protection and trust must be embedded throughout AI services," Kim said. "We will contribute to global standardization in AI digital trust, drawing on KT Corp's accumulated experience in personal data protection and its capabilities in international standardization."