Advised by the Ministry of Digital Affairs (MODA) and co-organized by the Administration for Digital Industries and the Taipei Computer Association (TCA), the Asia-Pacific flagship AI technology and application exhibition AI WAVE SHOW took place from July 31 to August 2 at the Taipei World Trade Center Hall 1. The Artificial Intelligence Evaluation Center (AIEC) participated in the “Future AI Pavilion” under the theme “AI Capability UP”, showcasing its five core evaluation dimensions—fairness, reliability, accuracy, privacy, and security—through immersive interactive games designed to raise public awareness of “trustworthy AI.” The center also presented its online automated evaluation simulation tool to help enterprises better understand the submission process and service workflow.

The exhibition was built around the concept “AIEC – Building a Trustworthy AI Partner,” offering tailored experiences for different target audiences. For the public, AIEC created a tarot-reading-style game that “explores the inner world of AI.” Each tarot card represented one of the evaluation dimensions; visitors drew cards and participated in Q&A sessions to learn how AI evaluations are conducted and to gain basic knowledge of trustworthy AI. For industry professionals, an online simulation demonstrated the end-to-end testing process—such as parameter settings and test item selection—using open-source language models as examples. This allowed participants to see how automated evaluations generate results, and how such results can guide model optimization and fine-tuning.

The booth attracted enthusiastic participation. Many visitors commented that the playful, hands-on approach helped them clearly understand what AI evaluation involves and what constitutes a “trustworthy AI.” Industry representatives—including those from the finance sector, AI startups, and ICT companies—expressed that evaluation is essential for AI development and that a neutral third-party verification system can be valuable for market readiness. Government representatives noted that such mechanisms could serve as a practical reference for AI governance and adoption.

Under MODA’s guidance, AIEC will continue to advance its AI evaluation mechanisms, focusing on developing more robust tools and methodologies grounded in international standards. The center aims to promote the globalization of Taiwan’s AI products, building an evaluation framework rooted in safety, stability, fairness, and transparency—realizing the vision of “Making the World Trust Taiwan’s AI.”