APMIC, a provider of enterprise-level AI solutions, announced today that its Traditional Chinese language model ACE-1-24B has officially passed the AIEC (Artificial Intelligence Evaluation Center) trustworthiness test, guided by the Ministry of Digital Development. This makes it one of the first Traditional Chinese models in Taiwan to pass this test, marking an important milestone in the development of Taiwan's AI industry.
Building trustworthy AI becomes a global trend
According to World Trade Organization data, while global enterprise AI adoption is projected to grow 115% between 2023 and 2024, only 62% of business leaders believe AI is being deployed responsibly. To strengthen Taiwan's AI competitiveness, the AIEC has established a comprehensive testing framework based on international standards such as NIST and ISO, covering key indicators such as security, fairness, and explainability.
APMIC Founder and CEO Wu Bohan said, "The future of AI models will be a race for standards and trust. The certification of ACE-1-24B signifies that traditional Chinese models have embarked on a standardized path that is both controllable and adaptable."
Solutions designed for high-standard industries
The ACE-1-24B is capable of processing 65KB of context. It targets training in three key areas: mathematical logic, everyday reasoning, and tool instructions. It supports chaining of thought and enhances the explainability of reasoning. Through model context protocols, it can connect to external systems for natural language-driven decision support.
It is specifically optimized for high-standard and complex fields such as finance, government, and manufacturing. It supports local deployment to ensure that data does not leave the country. It can handle applications such as regulatory inquiries, technical documents, and financial analysis, helping companies build trustworthy AI systems.
APMIC's passing of the AIEC evaluation demonstrates that Taiwan's AI development is shifting from a "technology race" to a "trust race." Compared to large international models, ACE-1-24B focuses on the traditional Chinese context and local compliance requirements, providing a more secure option for government and enterprise clients who prioritize data security.
As the Stanford report indicates that more than 40% of companies have improved their efficiency by introducing responsible AI, APMIC's move will pave the way for the implementation of AI in strictly regulated industries such as finance and government, and promote Taiwan's differentiated advantages in the global AI arena.
