Public and private sectors are joining forces to advance personal data protection systems in line with the rapid development of artificial intelligence (AI) technology.
The Personal Information Protection Commission held the 4th plenary meeting of the AI Privacy Public-Private Policy Council on July 8 at Seoul AI Hub in Seocho-gu, Seoul. The meeting brought together private experts and relevant government officials to intensively discuss personal data protection measures needed for the AI era.
Two key agenda items were addressed during the meeting: research on AI privacy risk diagnosis and certification plans, and guidelines for personal data processing in generative AI development and utilization.
Building AI Privacy Risk Diagnosis Framework
Professor Kim Byung-pil from KAIST (Chair of Policy Council’s 1st Working Group) presented specific diagnosis and certification plans for AI privacy risks. Based on comparative analysis of the latest technologies and best practices discussed in industry and academia, a framework for systematically diagnosing and evaluating privacy risks inherent in AI was shared.
The Personal Information Protection Commission had previously provided guidance on AI model privacy risk management procedures, risk types, and mitigation measures through an AI privacy risk management model released in December last year. Follow-up research is underway to develop systematic diagnosis and measurement standards to further advance this initiative.
Discussion on Generative AI Development Guidelines
Guidelines for personal data processing in generative AI development and utilization were also a major topic of discussion. These guidelines present legality and safety standards that must be considered for personal data processing and protection at each stage of the generative AI lifecycle.
The guidelines include ▲key provisions of personal data protection laws related to generative AI ▲AI personal data processing guidelines ▲enforcement cases including investigations and dispositions ▲innovative support cases such as regulatory sandboxes. They also reflect international trends by referencing the latest policy materials from major countries including the United States, United Kingdom, and European Union (EU).
Building Trust-Based Guardrails
Personal Information Protection Commission Chairman Go Hak-soo emphasized that “building safe guardrails is a prerequisite for trustworthy AI development,” adding that “privacy for AI innovation and AI innovation for privacy can serve as mechanisms to enhance technology acceptability and social trust through mutually beneficial relationships.”
The Personal Information Protection Commission plans to complete the guidelines for personal data processing in generative AI development and utilization by the end of July, incorporating opinions from civil society organizations, related ministries, and experts presented at this plenary meeting.
[Source: https://www.korea.kr/news/policyNewsView.do?newsId=148945645]
