Ministry of Justice Launches Sovereign AI Initiative with Specialized Large Language Model Project
Taiwan's Ministry of Justice held a press conference and forum on June 3 to launch its sovereign AI project, building a trustworthy, specialized large language model for legal affairs. Minister Cheng Ming-chien emphasized AI as an assistant, not a replacement. National Science and Technology Council Minister Wu Cheng-wen and Digital Affairs Minister Lin Yi-jing attended, aiming to establish a secure, transparent, and auditable government AI system.
📋 Article Processing Timeline
- 📰 Published: June 3, 2026 at 18:26
- 🔍 Collected: June 3, 2026 at 18:49 (23 min after Published)
- 🤖 AI Analyzed: June 6, 2026 at 16:34 (69h 44m after Collected)
(Central News Agency, Taipei, June 3) To implement the government's policy direction of promoting "Sovereign AI" and a smart technology island, the Ministry of Justice (MOJ) held a press conference and forum today to launch its specialized large language model for legal affairs. Minister of Justice Cheng Ming-chien stated that for the judiciary, AI must not only be intelligent but also trustworthy.
The MOJ held the "Building a Trustworthy Sovereign AI for Legal Affairs: Launch Press Conference and Forum for a Specialized Large Language Model," chaired by Minister Cheng. National Science and Technology Council (NSTC) Minister Wu Cheng-wen and Digital Affairs Minister Lin Yi-jing attended and delivered speeches. The MOJ also gathered representatives from industry, government, academia, research institutes, and its affiliated agencies, demonstrating its commitment to promoting trustworthy AI and strengthening judicial technological capabilities and digital governance.
In his speech, Minister Cheng highlighted the high demand for AI assistance in judicial practice areas such as fraud crimes, cross-border cases, and massive document processing. He emphasized that the MOJ's AI initiative is not about replacing legal expertise or judicial judgment with AI, but about using AI to help frontline colleagues reduce repetitive and time-consuming administrative work, allowing professional personnel to focus more on case analysis, risk identification, and public service.
Minister Cheng noted that introducing AI into the legal field involves people's rights and cannot solely pursue speed and efficiency. It must also ensure data security, result verifiability, and clear accountability. The AI system promoted by the MOJ will center on "trustworthy AI," implementing principles of privacy and information security, transparency and explainability, and fairness without bias. It will also introduce a third-party verification mechanism to ensure AI applications are built on a secure and trustworthy institutional foundation, with final judgment and responsibility remaining with humans.
Minister Wu Cheng-wen expressed hope that the project will continue to deepen three areas of work: continuously improving model capabilities and the depth of judicial applications to make AI a true smart assistant; continuously refining governance mechanisms for data, models, and computing power to establish standards and models for trustworthy government AI applications; and promoting cross-agency collaboration and the diffusion of results, transforming the experience accumulated in the legal field into an important foundation for other government agencies to develop sovereign AI.
Minister Lin Yi-jing stated that the judiciary inherently cannot tolerate ambiguity. The role of AI is not to "make decisions for you" but to "help organize information and improve efficiency, with every step being verifiable." This aligns with the concept of "trustworthy AI," which is not about how smart AI is, but whether it can be used with confidence in critical fields.
According to a press release from the MOJ, the Executive Yuan has been actively promoting the "New Ten Major AI Infrastructure Projects" in recent years, listing "Sovereign AI" as a key project for the digital foundation. For the highly sensitive field of judiciary, the MOJ recognizes the necessity of building an AI system that belongs to Taiwan and aligns with the values of democracy and the rule of law.
The press release stated that introducing AI into the legal field involves not only improving administrative efficiency but also core values such as the confidentiality of investigations, personal data protection, procedural justice, and the protection of people's rights. Therefore, the establishment of a specialized large language model for legal affairs is not just a digital construction project but a crucial undertaking to build a government AI system that is "governable, auditable, and trustworthy." (Editor: Li Hengshan) 1150603
The MOJ held the "Building a Trustworthy Sovereign AI for Legal Affairs: Launch Press Conference and Forum for a Specialized Large Language Model," chaired by Minister Cheng. National Science and Technology Council (NSTC) Minister Wu Cheng-wen and Digital Affairs Minister Lin Yi-jing attended and delivered speeches. The MOJ also gathered representatives from industry, government, academia, research institutes, and its affiliated agencies, demonstrating its commitment to promoting trustworthy AI and strengthening judicial technological capabilities and digital governance.
In his speech, Minister Cheng highlighted the high demand for AI assistance in judicial practice areas such as fraud crimes, cross-border cases, and massive document processing. He emphasized that the MOJ's AI initiative is not about replacing legal expertise or judicial judgment with AI, but about using AI to help frontline colleagues reduce repetitive and time-consuming administrative work, allowing professional personnel to focus more on case analysis, risk identification, and public service.
Minister Cheng noted that introducing AI into the legal field involves people's rights and cannot solely pursue speed and efficiency. It must also ensure data security, result verifiability, and clear accountability. The AI system promoted by the MOJ will center on "trustworthy AI," implementing principles of privacy and information security, transparency and explainability, and fairness without bias. It will also introduce a third-party verification mechanism to ensure AI applications are built on a secure and trustworthy institutional foundation, with final judgment and responsibility remaining with humans.
Minister Wu Cheng-wen expressed hope that the project will continue to deepen three areas of work: continuously improving model capabilities and the depth of judicial applications to make AI a true smart assistant; continuously refining governance mechanisms for data, models, and computing power to establish standards and models for trustworthy government AI applications; and promoting cross-agency collaboration and the diffusion of results, transforming the experience accumulated in the legal field into an important foundation for other government agencies to develop sovereign AI.
Minister Lin Yi-jing stated that the judiciary inherently cannot tolerate ambiguity. The role of AI is not to "make decisions for you" but to "help organize information and improve efficiency, with every step being verifiable." This aligns with the concept of "trustworthy AI," which is not about how smart AI is, but whether it can be used with confidence in critical fields.
According to a press release from the MOJ, the Executive Yuan has been actively promoting the "New Ten Major AI Infrastructure Projects" in recent years, listing "Sovereign AI" as a key project for the digital foundation. For the highly sensitive field of judiciary, the MOJ recognizes the necessity of building an AI system that belongs to Taiwan and aligns with the values of democracy and the rule of law.
The press release stated that introducing AI into the legal field involves not only improving administrative efficiency but also core values such as the confidentiality of investigations, personal data protection, procedural justice, and the protection of people's rights. Therefore, the establishment of a specialized large language model for legal affairs is not just a digital construction project but a crucial undertaking to build a government AI system that is "governable, auditable, and trustworthy." (Editor: Li Hengshan) 1150603
FAQ
What is the purpose of this project?
To promote AI adoption in Taiwan's judicial field, improve administrative efficiency, and build a trustworthy government AI system based on democratic and rule-of-law values.
Who is leading this project?
The Ministry of Justice is leading the project, with collaboration from the National Science and Technology Council and the Ministry of Digital Affairs.
When was it launched?
The launch was officially announced at a press conference and forum on June 3, 2025.