Seeking Practical Ways to Secure 'Transparency' in AI System Development

2024.08.01. The Fair Economy — The Trustworthy AI International Network (TRAIN) announced that it will hold a seminar on "artificial intelligence (AI) trustworthiness" on the 7th at Entreville in Gangnam-gu, Seoul.

According to TRAIN, this event is the first in a newly launched regular seminar series, and TRAIN plans to hold six AI trustworthiness seminars by the end of this year.

The first event will focus on practical ways to secure "transparency" needed in the AI system development process, based on the latest laws and standards. It will then go on to address data governance and bias, and the trustworthiness of AI models and systems in turn.

Experts from various institutions and companies, including Won Law Firm, the Korea Testing Laboratory (KTL), and ThinkforBL, will give presentations and lead discussions in the fields of law, standards, and technology, respectively. They will introduce what transparency is and the latest cases and information that can be practically applied to secure such transparency in the field of technology development.

Ko In-seon, a lawyer on the AI Response Team at Law Firm Won, will be the first speaker. Under the theme "AI Transparency Required by Law," she will define what legal considerations are necessary to ensure AI transparency.

Ahn Seon-ho, head of KTL's AI team and also a member of the SC42 subcommittee of the AI International Standardization Plenary (ISO/IEC JTC1/SC42), will present how transparency should be addressed from a technical perspective under the topic "Analysis of the Definition and Framework of International Standard AI Transparency."

Cheon Seon-il, a senior researcher at ThinkforBL who oversaw the production of a guide for developing trustworthy AI, will discuss "International Cases of Applying AI Transparency." He will compare and present transparency requirements needed in AI system development along with domestic and international application cases.

TRAIN stated, "This seminar is meaningful in that it seeks practical ways to respond to the rapidly changing global AI technology regulation trends. It will provide differentiated value compared to other seminars covering similar topics. We have prepared the event to proceed in a free atmosphere, allowing two-way communication with the audience, not just one-way information delivery by the speakers."

It added, "We have thoroughly prepared the event from a practical perspective so that practitioners struggling with regulatory responses in the AI technology field can understand and approach the topics relatively easily. In addition to its consulting, education, and certification businesses, TRAIN plans to focus on delivering useful information on the ground by holding regular seminars as well as national touring lectures for administrative practitioners."

Meanwhile, this seminar, hosted by TRAIN and co-organized by ThinkforBL and Won Law Firm, is sponsored by the Korea Testing Laboratory (KTL), the Jeonju Information & Culture Industry Promotion Agency (JICA), the Busan IT Convergence Components Research Institute (CIDI), the Korea Software Industry Association (KOSA), Koiia, the International AI Ethics Association (IAAE), and CATBELL Company.