Tscientific Wins KISA National Project... "Advancing AI-Based Personal Information Protection Technology"

Tscientific announced on April 28 that a consortium led by Tscientific, with WorldVertech and the Yonsei University-Industry Cooperation Foundation as participants, has been selected as the final research and development institution for the "Development of Multimodal Artificial Intelligence (AI)-Based Personal Information Detection, Tracking, and De-identification Technology" project, overseen by the Personal Information Protection Commission and the Korea Internet & Security Agency (KISA).

The goal of this project is to develop technology that can use AI to detect, track, and de-identify personal information from multimodal data, which consists of both structured and unstructured data such as text, images, audio, and video.

The project is scheduled to run until 2027 and aims to develop next-generation technology that uses AI to detect and protect personal information contained in multimodal data generated during the use of generative AI services.

The main technologies to be developed include: ▲ technology for context-aware detection and tracking of personal information located in unspecified positions within multimodal data; ▲ technology for de-identifying personal information while preserving key information in the original data; ▲ and the construction of personal information detection and de-identification datasets, as well as the development of a system framework.

Lee Jongkyu, head of the Tscientific Research Institute, stated, "Being selected as the research and development institution for this project is further recognition of our company’s excellence in AI-based personal information protection technology." He added, "Based on the achievements of our previous project, 'Development of AI-Based Personal Information Detection and De-identification Technology in Conversational Text Data,' conducted with WorldVertech and Yonsei University, we will expand our scope to multimodal data and develop even more advanced technology."

He continued, "This technology development initiative is being pursued as a proactive government response to the increased risk of personal information leakage and infringement resulting from the emergence of generative multimodal AI models." He also said, "We expect that this will contribute to improving the quality of safe and valuable AI training data and vitalizing the data industry ecosystem by protecting against risks such as personal information infringement or leakage."

Meanwhile, Tscientific previously obtained Grade 1 GS (Good Software) certification last year for its AI technology-based unstructured personal information detection solution, 'P3R-Privacy Finder AI.' The company has also continued to achieve results in the field of personal information protection technology, such as publishing papers on 'Personal Information Protection Using Artificial Intelligence' with the research team of Professor Kim Hansem at Yonsei University.