Artificial intelligence (AI) technology startup Tunip (CEO Park Gyu-byung) announced that it has been selected for the "SCALE-UP TIPS" program, a private investment-led technology startup support program hosted by the Ministry of SMEs and Startups. DSC Investment participated as the operating company for this project.
Scaleup Tips is a program where the government and private sector collaborate to identify promising startups and support their technological advancement and global growth. Tunip, upon selection, will receive a total of 1.2 billion won in R&D and commercialization funding over the next three years.
Tunip is a company that develops AI chatbots and AI ethics guardrail solutions based on large-scale Korean-specific language models and speech and morphological analysis technologies. In particular, the company is focusing on developing its "AI ethics guardrail" technology, which detects and blocks harmful generative AI speech, personal information disclosure, and prompt attacks, thereby building a safe AI ecosystem.
Through this Scale-Up TIPS project, Tunip will develop a Korean-specific multimodal guardrail system. The system aims to secure technology that comprehensively detects sensitive topics and hate speech in text, voice, image, and video environments, protects personal information, prevents prompt attacks, and controls responses based on regulatory standards.
Additionally, we plan to support the introduction of AI agent security in various industries such as finance, public sector, education, and commerce by providing lightweight model optimization technology and API/SDK-based developer tools that can be used in on-premises, cloud, and mobile environments.
Park Gyu-byeong, CEO of Tunip, said that in the era of AI agents, guardrail technology that guarantees safety and reliability as well as performance will become the core of industrial competitiveness, and expressed his intention to strengthen Korean-specific multimodal guardrail technology through Scale Up Tips and contribute to establishing global AI safety standards.
- See more related articles
You must be logged in to post a comment.