Research on Taiwan’s Policies of Innovative Industry Development in Recent Years (2015-2016)
1. “Five plus Two” Innovative Industries Policy
On June 15, 2016, Premier Lin Chuan met with a group of prominent business leaders to talk about a government project on five innovative industries, which aim to drive the next generation of businesses in R.O.C.. Subsequently the program was expanded to include “new agriculture” and the “circular economy” as the “+2.” The program was then broadened even further to include the Digital Economy and Cultural Innovation, with even Semiconductors and IC Design included, although the name of the policy remains 5+2. Speaking at the Third Wednesday Club in Taipei, Premier Lin said the industries require more investment to drive the next generation of industry growth momentum in R.O.C., create high-quality jobs, and upgrade the industrial competitiveness. Executive Yuan has selected the five innovative industries of Asia Silicon Valley, smart machinery, green energy, biotech & pharmaceutical industry, and national defense, which will be the core for pushing forward the next-generation industrial growth and improve overall environment by creating a cluster effect that links local and global industries, while simultaneously raising wages and stimulating employment.
Premier Lin said, regarding industrial competitiveness and investment issues the lackluster economy has stifled investment opportunities, and with limited government budgets, the private sector must play the larger role in investments. Regarding the “Five major Innovative Industries” project, Premier Lin said the National Development Council is currently drafting long-term plan to attract talent, create a thriving working environment, and infuse companies with more innovation, entrepreneurship and young workers. In addition, R.O.C. must also cultivate a strong software industry, without which it would be difficult to build a highly intelligent infrastructure.
The National Development Council said the program possess both the capacity of domestic demand and local characteristics, as the core for pushing forward the next-generation industrial growth. The government aims to promote a seamless synergy of investment, technology, and the talent, in order to develop innovative industrial clusters for furthering global linkage and nurturing international enterprises. In the meantime, the government also aims at achieving the enhancement of technology levels, balanced regional development, as well as realizing the benefits of job creation.
2. The Asia Silicon Valley Development Plan
In September 2016 the government approved the Asia Silicon Valley Development Plan, which connect Taiwan to global tech clusters and create new industries for the next generation. By harnessing advanced technological research and development results from around the world, the plan hopes to promote innovation and R&D for devices and applications of the internet of things (IoT), and upgrade Taiwan’s startup and entrepreneurship ecosystem.
The four implementation strategies are as follows:
(1) Building a comprehensive ecosystem to support innovation and entrepreneurship
(2) Connect with international research and development capabilities
(3) Create an IoT value chain
(4) Construct diversified test beds for smart products and services by establishing a quality internet environment
Taiwan’s first wave of industrial development was driven by continuous technological innovation, and the wave that followed saw the information industry become a major source of economic growth.
3. Global Hub for Smart Machinery
On July 21, 2016, Premier Lin Chuan said at a Cabinet meeting, the government aims to forge Taiwan into a global manufacturing hub for intelligent machinery and high-end equipment parts. Upgrading from precision machinery to intelligent machinery is the main goal of putting intelligent machinery industry into focal execution area expecting to create jobs and to maximize the production of production line as well as to forge central Taiwan into a global manufacturing hub for smart machinery. The Ministry of Economic draws up the Intelligent Machinery Promotion Program to establish the applications of the technology and capacity of services that fit the demand of the market. The program embodies two parts. The first is to accelerate the industrialization of intelligent machinery for building an ecosystem. The second is to improve intelligentization by means of introducing the intelligent machinery into the industries.
The execution policy of the Intelligent Machinery Promotion Program is to integrate the intelligent functions such as malfunctions predictions, accuracy compensation, and automatic parameter setting into the machinery industry so as to have the ability to render the whole solutions to the problem. Simultaneously, the program employs three strategies, which are connecting with the local industries, connecting with the future, and connecting with the world, to develop the mentioned vision and objectives. Especially, the way to execute the strategy of connecting with the local industries consists of integrating the capabilities of industry, research organization and the government. At the meantime, the government will encourage the applications of smart vehicles and unmanned aerial vehicles and train the talents as well. The thinking of connecting with the future lies in the goal of deepening the technologies, establishing systematic solutions, and providing a testing areas, which focus on the related applications such as aerospace, advanced semiconductor, smart transportation, green vehicles, energy industry, whole solutions between factories, intelligent man-machine coordination, and robots of machine vision combined with intelligent machinery applications. The government would strengthen the cross-cutting cooperation to develop machines for aerospace and integrate the system of industrial division to form a cluster in order to create Taiwanese IoT technology. Eventually, Taiwan will be able to connect with the world, enhance international cooperation, expand export trade and push industry moving toward the age of information and digital economy and break the edge of industry technology to make the industry feel the goodwill of the government.
4. Green energy innovations
The government’s “five plus two” innovative industries program includes a green energy industrial innovation plan passed October 27, 2016 that will focus on Taiwan’s green needs, spur extensive investments from within and outside the country, and increase quality employment opportunities while supporting the growth of green energy technologies and businesses.
The government is developing the Shalun Green Energy Science City. The hub’s core in Shalun will house a green energy technology research center as well as a demo site, providing facilities to develop research and development (R&D) capabilities and conduct the requisite certification and demonstration procedures. The joint research center for green energy technologies will integrate the efforts of domestic academic institutions, research institutes, state-run enterprises and industry to develop green energy technologies, focusing on four major functions: creating, conserving and storing energy, as well as system integration. Development strategies include systems integration and finding better ways to conserve, generate and store energy by promoting green energy infrastructure, expanding renewable energy capabilities and cooperating with large international firms.
The emergence of the green economy has prompted the government to build infrastructure that will lay the foundation for Taiwan’s green energy sector, transform the nation into a nuclear-free society, and spur industrial innovation. For innovative technology industries, green energy industries can drive domestic economic development by attracting more venture capital and creating more employment opportunities.
5. Biomedical Industry Innovation Program
To facilitate development of Taiwan’s biomedical industry, the government proposed a “biomedical industrial innovation promotion program” on November 10, 2016 to serve as the nation’s new blueprint for innovative biomedical research and development (R&D). To facilitate development of the biomedical industry, the government proposed a “biomedical industrial innovation promotion program”. The program centered on the theme of “local, global and future links,” “the biomedical industrial innovation promotion program” includes four action plans:
(1) Build a comprehensive ecosystem
To address a rapidly ageing global population, Taiwan will enhance the biomedical industry’s capacity for innovation by focusing on talent, capital, topic selection, intellectual property, laws and regulations, and resources.
(2) Integrate innovative business clusters
Established by the Ministry of Science and Technology and based in Hsinchu Biomedical Science Park, the center will serve as a government think tank on related issues. It is also tasked with initiating and advancing exchanges among local and foreign experts, overseeing project implementation, promoting investment and recruiting talents. Equally important, it will play a central role in integrating resources from other biomedical industry clusters around the country, including Nangang Software Park in Taipei City, Central Taiwan Science Park in Taichung City and Southern Science Park in Tainan City.
(3) Connect global market resources
Building on Taiwan’s advantages, promote M&A and strategic alliances, and employ buyout funds and syndicated loans to purchase high-potential small and medium-sized international pharmaceutical companies, medical supply companies, distributors and service providers. Use modern mosquito-borne disease control strategies as the foundation of diplomatic cooperation, and promote the development of Taiwan’s public health care and medical services in Southeast Asian countries.
(4) Promote specialized key industries
Promote niche precision medical services, foster clusters of world-class specialty clinics, and develop industries in the health and wellness sectors.
6. DIGITAL NATION AND INNOVATIVE ECONOMIC DEVELOPMENT PLAN
On November 24, 2016, the Executive Yuan promote the Digital Nation and Innovative Economic Development Plan (2017-2025) (DIGI+ program), the plan’s main goals for 2025 are to grow R.O.C.’s digital economy to NT $ 6.5 trillion (US$205.9 billion), increase the digital lifestyle services penetration rate to 80 percent, speed up broadband connections to 2 Gbps, ensure citizens’ basic rights to have 25 Mbps broadband access, and put R.O.C. among the top 10 information technology nations worldwide.
In addition to the industrial economy, the program can jump off bottlenecks in the past industrial development, and promote the current Internet of things, intelligent machinery, green energy, medical care and other key national industries, but also attaches great importance to strengthening the digital infrastructure construction, the development of equal active, as well as the creation of a service-oriented digital government. It is also hoped that through the construction of a sustainable and intelligent urban and rural area, the quality of life will be improved and the people will enjoy a wealthy and healthy life. Over the next 8 years, the government will spend more than NT $ 150 billion.
The plan contains several important development strategies: DIGI+Infrastructure: Build infrastructure conducive to digital innovation. DIGI+Talent: Cultivate digital innovation talent. DIGI+Industry: Support cross-industry transformation through digital innovation. DIGI+Rights: Make R.O.C. an advanced society that respects digital rights and supports open online communities. DIGI+Cities: Build smart cities through cooperation among central and local governments and the industrial, academic and research sectors. DIGI+Globalization: Boost R.O.C.’s standing in the global digital service economy.
The program aims to build a favorable environment for digital innovation and to create a friendly legal environment to complete the draft amendments to the Digital Communications Law and the Telecommunications Act as soon as possible, foster cross-domain digital talents and develop advanced digital technologies, To create a digital economy, digital government, network society, smart urban and rural and other national innovation ecological environment in order to achieve "the development of active network society, promote high value innovation economy, open up rich countries of the policy vision.
In order to achieve the overall effectiveness of the DIGI + program, interdisciplinary, inter-ministerial, inter-departmental and inter-departmental efforts will be required to collaborate with the newly launched Digital National Innovation Economy (DIGI +) Promotion Team.
7. “NEW AGRICULTURE” PROMOTION PROJECT
At a Cabinet meeting On December 08, 2016, Premier Lin Chuan underscored the importance of a new agricultural paradigm for Taiwan’s economic development, adding that new agriculture is an integral part of the “five plus two” industrial innovation projects proposed by President Tsai Ing-wen. The “new agriculture” promotion project uses innovation technology to bring value to agricultural, and build new agricultural paradigm, agricultural safety systems and promote agricultural marketing. This project also takes resources recycling and environmental sustainability into consideration to promote agricultural transformation, and build a robust new agricultural system.
This agricultural project is expected to increase food self-sufficiency rate to 40%, level up agricultural industry value by NT$43.4 billion, create 370,000 jobs and increase portion of total agricultural exports to new overseas markets to 57% by 2020.
This project contains three aspects:
First is “building new agricultural paradigm”: to protect farmers, agricultural development and ensure sustainability of the environment.
Second is “building agricultural safety systems”: Ensuring product safety and quality, and building a certification system which can be trust by the consumers and is consistent with international standards.
Last but not least is “leveling up agricultural marketing and promotions”: enhancing promotion, making the agricultural industry become profitable and sustainable.
Council of Agriculture’s initiatives also proposed 10 policies to leverage agricultural industry, not only just use the passive subsidies measure of the past. These policies including promoting environmentally friendly farming practices; giving farmers that are beneficial(green) to the land payments; stabilizing farmers’ incomes; increasing the competitiveness of the livestock and poultry industries; using agricultural resources sustainably; ensuring the safety of agricultural products; developing technological innovation; leveling up food security; increasing diversification of domestic and external marketing channels; and increasing agriculture industry added value.
In this statutes report, Council of Agriculture said this project will accelerate reforms, create new agricultural models and safety systems, but also build a new sustainable paradigm of agricultural. Premier Lin Chuan also backed this “five plus two innovative industries” program and “new agriculture” project, and asked Council of Agriculture to reviewing the possible legal changes or amendment that may help to enhance the transformation of agricultural sector.
Blockchain and General Data Protection Regulation (GDPR) compliance issues (2019) I. Brief Blockchain technology can solve the problem of trust between data demanders and data providers. In other words, in a centralized mode, data demanders can only choose to believe that the centralized platform will not contain the false information. However, in the decentralized mode, data isn’t controlled by one individual group or organization[1], data demanders can directly verify information such as data source, time, and authorization on the blockchain without worrying about the correctness and authenticity of the data. Take the “immutable” for example, it is conflict with the right to erase (also known as the right to be forgotten) in the GDPR.With encryption and one-time pad (OTP) technology, data subjects can make data off-chain storaged or modified at any time in a decentralized platform, so the problem that data on blockchain not meet the GDPR regulation has gradually faded away. II. What is GDPR? The purpose of the EU GDPR is to protect user’s data and to prevent large-scale online platforms or large enterprises from collecting or using user’s data without their permission. Violators will be punished by the EU with up to 20 million Euros (equal to 700 million NT dollars) or 4% of the worldwide annual revenue of the prior financial year. The aim is to promote free movement of personal data within the European Union, while maintaining adequate level of data protection. It is a technology-neutral law, any type of technology which is for processing personal data is applicable. So problem about whether the data on blockchain fits GDPR regulation has raise. Since the blockchain is decentralized, one of the original design goals is to avoid a large amount of centralized data being abused. Blockchain can be divided into permissioned blockchains and permissionless blockchains. The former can also be called “private chains” or “alliance chains” or “enterprise chains”, that means no one can join the blockchain without consent. The latter can also be called “public chains”, which means that anyone can participate on chain without obtaining consent. Sometimes, private chain is not completely decentralized. The demand for the use of blockchain has developed a hybrid of two types of blockchain, called “alliance chain”, which not only maintains the privacy of the private chain, but also maintains the characteristics of public chains. The information on the alliance chain will be open and transparent, and it is in conflict with the application of GDPR. III. How to GDPR apply to blockchain ? First, it should be determined whether the data on the blockchain is personal data protected by GDPR. Second, what is the relationship and respective responsibilities of the data subject, data controller, and data processor? Finally, we discuss the common technical characteristics of blockchain and how it is applicable to GDPR. 1. Data on the blockchain is personal data protected by GDPR? First of all, starting from the technical characteristics of the blockchain, blockchain technology is commonly decentralized, anonymous, immutable, trackable and encrypted. The other five major characteristics are immutability, authenticity, transparency, uniqueness, and collective consensus. Further, the blockchain is an open, decentralized ledger technology that can effectively verify and permanently store transactions between two parties, and can be proved. It is a distributed database, all users on the chain can access to the database and the history record, also can directly verify transaction records. Each nodes use peer-to-peer transmission for upload or transfer information without third-party intermediation, which is the unique “decentralization” feature of the blockchain. In addition, the node or any user on the chain has a unique and identifiable set of more than 30 alphanumeric addresses, but the user may choose to be anonymous or provide identification, which is also a feature of transparency with pseudonymity[2]; Data on blockchain is irreversibility of records. Once the transaction is recorded and updated on the chain, it is difficult to change and is permanently stored in the database, that is to say, it has the characteristics of “tamper-resistance”[3]. According to Article 4 (1) of the GDPR, “personal data” means any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person. Therefore, if data subject cannot be identified by the personal data on the blockchain, that is an anonymous data, excluding the application of GDPR. (1) What is Anonymization? According to Opinion 05/2014 on Anonymization Techniques by Article 29 Data Protection Working Party of the European Union, “anonymization” is a technique applied to personal data in order to achieve irreversible de-identification[4]. And it also said the “Hash function” of blockchain is a pseudonymization technology, the personal data is possible to be re-identified. Therefore it’s not an “anonymization”, the data on the blockchain may still be the personal data stipulated by the GDPR. As the blockchain evolves, it will be possible to develop technologies that are not regulated by GDPR, such as part of the encryption process, which will be able to pass the court or European data protection authorities requirement of anonymization. There are also many compliance solutions which use technical in the industry, such as avoiding transaction data stored directly on the chain. 2. International data transmission Furthermore, in accordance with Article 3 of the GDPR, “This Regulation applies to the processing of personal data in the context of the activities of an establishment of a controller or a processor in the Union, regardless of whether the processing takes place in the Union or not. This Regulation applies to the processing of personal data of data subjects who are in the Union by a controller or processor not established in the Union, where the processing activities are related to: (a) the offering of goods or services, irrespective of whether a payment of the data subject is required, to such data subjects in the Union; or (b) the monitoring of their behaviour as far as their behaviour takes place within the Union”.[5] In other words, GDPR applies only when the data on the blockchain is not anonymized, and involves the processing of personal data of EU citizens. 3. Identification of data controllers and data processors Therefore, if the encryption technology involves the public storage of EU citizens' personal data and passes it to a third-party controller, it may be identified as the “data controller” under Article 4 of GDPR, and all nodes and miners of the platform may be deemed as the “co-controller” of the data, and be assumed joint responsibility with the data controller by GDPR. For example, the parties can claim the right to delete data from the data controller. In addition, a blockchain operator may be identified as a “processor”, for example, Backend as a Service (BaaS) products, the third parties provide network infrastructure for users, and let users manage and store personal data. Such Cloud Services Companies provide online services on behalf of customers, do not act as “data controllers”. Some commentators believe that in the case of private chains or alliance chains, such as land records transmission, inter-bank customer information sharing, etc., compared to public chain applications: such as cryptocurrencies (Bitcoin for example), is not completely decentralized, and more likely to meet GDPR requirements[6]. For example, in the case of a private chain or alliance chain, it is a closed platform, which contains only a small number of trusted nodes, is more effective in complying with the GDPR rules. 4. Data subject claims In accordance with Article 17 of the GDPR, The data subject shall have the right to obtain from the controller the erasure of personal data concerning him or her without undue delay and the controller shall have the obligation to erase personal data without undue delay under some grounds. Off-chain storage technology can help the blockchain industry comply with GDPR rules, allowing offline storage of personal data, or allow trusted nodes to delete the private key of encrypted information, which leaving data that cannot be read and identified on the chain. If the data is in accordance with the definition of anonymization by GDPR, there is no room for GDPR to be applied. IV. Conclusion In summary, it’s seem that the application of blockchain to GDPR may include: (a) being difficulty to identified the data controllers and data processors after the data subject upload their data. (b) the nature of decentralized storage is transnational storage, and Whether the country where the node is located, is meets the “adequacy decision” of Article 45 of the GDPR. If it cannot be met, then it needs to consider whether it conforms to the transfers subject to appropriate safeguards of Article 46, or the derogations for specific situations of Article 49 of the GDPR. Reference: [1] How to Trade Cryptocurrency: A Guide for (Future) Millionaires, https://wikijob.com/trading/cryptocurrency/how-to-trade-cryptocurrency [2] DONNA K. HAMMAKER, HEALTH RECORDS AND THE LAW 392 (5TH ED. 2018). [3] Iansiti, Marco, and Karim R. Lakhani, The Truth about Blockchain, Harvard Business Review 95, no. 1 (January-February 2017): 118-125, available at https://hbr.org/2017/01/the-truth-about-blockchain [4] Article 29 Data Protection Working Party, Opinion 05/2014 on Anonymisation Techniques (2014), https://www.pdpjournals.com/docs/88197.pdf [5] Directive 95/46/EC (General Data Protection Regulation), https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32016R0679&from=EN [6] Queen Mary University of London, Are blockchains compatible with data privacy law? https://www.qmul.ac.uk/media/news/2018/hss/are-blockchains-compatible-with-data-privacy-law.html
A Preferred Model for Taiwan’s agency level AI risk categorization and management: A Cross-Jurisdictional PerspectiveA Preferred Model for Taiwan’s agency level AI risk categorization and management: A Cross-Jurisdictional Perspective 2025/09/15 Taiwan’s draft Artificial Intelligence Basic Law includes a provision allowing each government agency to establish its own risk-based AI management rules tailored to sector-specific regulatory needs[1]. To strike an effective balance between innovation and oversight, selecting an appropriate reference model is essential. After comparing major jurisdictions, this research argues that the United States Office of Management and Budget (OMB) Memorandum M-25-21—Accelerating Federal Use of AI through Innovation, Governance, and Public Trust[2]—offers the most balanced and practical approach for Taiwan’s agencies to refer to at this initial stage of developing AI regulation and promoting AI adoption. This article will first present an overview of the U.S. M-25-21 framework and its key features. It will then explain why the U.S. model is more suitable for Taiwan than those of other jurisdictions. Finally, it will conclude with recommendations for the government. I. Overview of the U.S. M-25-21 Framework Issued in April 2025 under Executive Order 14179, M-25-21 directs federal agencies to accelerate the adoption of artificial intelligence while maintaining a set of minimum safeguards. The memorandum identifies three priorities—innovation, governance, and public trust—and structures AI oversight around these principles. It requires every executive branch agency to designate a Chief AI Officer (CAIO), a senior official empowered to promote AI innovation, maintain a current inventory of AI use cases, and ensure that processes such as determining “high-impact” uses are in place. Rather than imposing a centralized management system, M-25-21 allows each agency to make context-sensitive determinations and to accept or waive risk management requirements. This approach recognizes that agencies vary widely in mission and capacity and are best positioned to understand the potential risks and benefits of AI within their own domains. The memorandum defines high-impact AI as systems whose outputs serve as a principal basis for decisions or actions with legal, material, binding, or significant rights and safety consequences. It offers a non-exhaustive list of presumed high-impact categories, including safety-critical functions of critical infrastructure, traffic management, patient diagnosis, blocking protected speech, and law enforcement applications. If an agency official determines that a specific AI use within these categories does not meet the high-impact definition, they must submit written documentation to notify the CAIO. By tying the definition to the effect of an AI system’s output rather than to a fixed sectoral list, M-25-21 provides a flexible method for identifying high-risk AI applications while preserving room for innovation. II. Key Features of the U.S. M-25-21 Framework A. Minimum Risk Management Practices To ensure protection without creating excessive barriers, M-25-21 specifies a set of minimum risk management practices that each agency must apply when using high-impact AI. Agencies are required to conduct pre-deployment testing under realistic conditions to confirm that AI systems perform as intended and to prepare appropriate risk mitigation plans. Even when agencies lack access to source code or training data, they are expected to use alternative testing methods—such as querying the AI service and observing its outputs—to assess performance and potential risks. Before deploying a high-impact AI system, agencies must complete an AI impact assessment. This assessment must explain the system’s intended purpose and expected benefits, analyze the quality and appropriateness of the data used, and evaluate potential impacts on privacy, civil rights, and civil liberties. It should also include a cost analysis, planned reassessment schedules and procedures, and comments highlighting potential concerns or gaps from an independent reviewer who was not involved in the system’s development. Importantly, the assessment must carry the signature of an accountable official who formally accepts the risk of deploying the AI system. Once deployed, agencies are expected to monitor AI systems continuously for performance drift, security vulnerabilities, or unforeseen adverse effects, and to implement appropriate mitigations and maintain documentation. Human oversight is equally essential: operators must receive specific training to interpret AI outputs, intervene when necessary, and use fail-safes or override mechanisms to minimize the risk of significant harm in high-impact situations. To protect the public, M-25-21 insists that individuals affected by AI-enabled decisions have access to timely human review and opportunities to appeal adverse outcomes. Appeals should not impose unnecessary burdens on individuals or the administration. Furthermore, agencies are expected to seek feedback from end users and the public to inform AI-related decision-making. These combined practices—testing, assessment, independent review, monitoring, human oversight, remedies, and feedback—form a balanced foundation for responsible AI use. The memorandum also requires agencies to safely discontinue any high-impact use cases that fail to comply with the minimum practices. B. Waiver System: Purpose and Conditions A distinctive feature of M-25-21 is its formal system for waivers from the minimum risk management practices. The waiver mechanism exists to reconcile two priorities: ensuring safety and rights protections on the one hand, and enabling innovation and rapid response on the other. Waivers may be considered when following a particular requirement would actually increase risks to safety or rights overall, or when compliance would create an unacceptable impediment to critical agency operations. For example, during a natural disaster or public health emergency, strict adherence to every procedural requirement might delay the deployment of an AI application that could save lives. In such situations, the CAIO may authorize a waiver to permit rapid deployment while still tracking and reassessing the use. Waivers for pilot programs are equally important for encouraging experimentation and innovation. They allow agencies to conduct small-scale, time-limited AI projects without implementing all minimum risk management practices, provided certain conditions are met: the pilot must be certified by the CAIO, centrally tracked, offer opt-in and opt-out options for individual participation, and apply minimum risk management practices where practicable. The memorandum imposes safeguards on this flexibility. Every waiver must be documented with a written determination explaining the reasoning, centrally tracked, and reassessed annually or whenever significant changes to the AI application’s conditions or context occur. CAIOs retain the power to revoke waivers at any time, and agencies must report any granted or revoked waiver to OMB annually and within 30 days of significant modifications. This approach maintains accountability while preventing rigid rules from becoming obstacles to effective governance. C. Disclosure Requirements for High-Impact Use and Waivers M-25-21 strongly emphasizes transparency as a pillar of public trust. Each agency must maintain an inventory of all AI use cases, submit it to OMB, and post a public version on the agency’s website. This inventory should be updated annually and, ideally, throughout the year to reflect the agency’s current use of AI. Transparency ensures that the public, civil society, and oversight bodies can understand where AI is influencing important government decisions without exposing sensitive or classified details. Similarly, agencies must publicly release summaries of each waiver or determination, including the justification, or explicitly indicate when no determinations or waivers are active. By making these summaries visible, the system builds confidence that waivers are granted for legitimate reasons. At the same time, OMB retains the authority to request detailed records concerning exception determinations within presumed high-impact categories. This combination of public disclosure and federal oversight helps maintain trust while safeguarding privacy, national security, and proprietary information. III. Why M-25-21 Stands Out for Taiwan’s AI Governance among Global Approaches Taiwan’s draft AI Basic Law envisions a decentralized system in which each agency determines its own risk classification and management practices[3]. The U.S. framework aligns closely with this philosophy. By empowering agencies to identify high-risk AI use cases tailored to their specific contexts, M-25-21 helps ensure that AI governance remains grounded in operational realities. At the same time, adopting M-25-21’s baseline practices, waiver safeguards, and disclosure requirements would provide consistency and public accountability across agencies. The combination of minimum risk management practices and transparent waiver use would encourage innovation while reassuring the public that any exceptions are justified, continuously monitored, and effectively controlled. Furthermore, embracing an approach that reflects emerging international consensus—particularly the emphasis on transparency in both U.S. and EU regimes—would position Taiwan to harmonize with global AI governance trends and strengthen its credibility in international markets. In contrast, the European Union’s AI Act predefines high-risk categories and mandates strict conformity assessments, CE Marking, and post-market monitoring[4]—an approach that is comprehensive but resource-intensive and may not suit all agencies equally. Australia’s ongoing discussions had been trending toward a similarly comprehensive model, but there has recently been backlash against this approach. Korea’s AI Basic Act[5] references high-risk AI only in broad terms and leaves most operational details undefined. M-25-21 strikes a middle ground, offering minimum yet concrete safeguards while preserving the flexibility agencies need to tailor governance to their specific domains. IV. Recommendations and Conclusion Based on this analysis, this research recommends that each agency designate a senior AI leader similar to a CAIO, maintain a public inventory of high-impact AI use cases, and publish summaries of waivers or determinations while safeguarding sensitive information. Agencies should also be encouraged to share AI resources and lessons learned to reduce duplication and strengthen governance maturity across government. Over time, these risk management practices can be refined in response to operational experience and evolving international standards. By adopting these principles, Taiwan can empower its agencies to innovate responsibly, protect citizens’ rights, and build public trust—ensuring that AI deployment across government remains both effective and aligned with global best practices. [1]〈政院通過「人工智慧基本法」草案 建構AI發展與應用良善環境 打造臺灣成為AI人工智慧島〉,行政院,https://www.ey.gov.tw/Page/9277F759E41CCD91/5d673d1e-f418-47dc-ab35-a06600f77f07(最後瀏覽日期︰2025/09/15)。 [2] United States Office of Management and Budget (OMB), M-25-21 Accelerating Federal Use of AI through Innovation, Governance, and Public Trust, https://www.whitehouse.gov/wp-content/uploads/2025/02/M-25-21-Accelerating-Federal-Use-of-AI-through-Innovation-Governance-and-Public-Trust.pdf (last visited Sept 15, 2025). [3] 蘇文彬,〈行政院通過AI基本法草案,將不設立AI專責機關〉,iThome,https://www.ithome.com.tw/news/170874(最後瀏覽日期︰2025/09/15)。 [4] Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act), https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32024R1689 (last visited Sept 15, 2025). [5] 인공지능발전과신뢰기반조성등에관한기본법안,https://www.law.go.kr/%EB%B2%95%EB%A0%B9/%EC%9D%B8%EA%B3%B5%EC%A7%80%EB%8A%A5%20%EB%B0 %9C%EC%A0%84%EA%B3%BC%20%EC%8B%A0%EB%A2%B0%20%EA%B8%B0%EB%B0%98%20%EC%A1%B0 %EC%84%B1%20%EB%93%B1%EC%97%90%20%EA%B4%80%ED%95%9C%20%EA%B8%B0%EB%B3%B8%EB%B2 %95/(20676,20250121) (last visited Sept 15, 2025).
The Status of Taiwan's Regulations Concerning with Access to Biological ResourcesPreface In actual practice, the research and development of biotech medicine, food, and environmental products cannot be done by in-lab researches. This is a unique character of the biotechnology industry. To get the research going, the researchers need to search for and exploit new biological materials and, samples outside the lab. Therefore, the access to and management of biological resources have significant impact on the stimulation and development of national biotech industry. Ever since the enforcement of Convention on Biological Diversity (CBD) in 1992 by 172 countries, a general principal about conserving biological diversity and using biological resources has been set. According to CBD, States have sovereign rights over their own biological resources. CBD also encourages each State to access to and manage the biological resources conformed with the principals of conservation, sustainability, NOEL environment friendly, and adequate sharing of benefit arising from biological resources. Therefore, issues such as environmental protection and sustainability have become political issues internationally. If the ABS system for the access to biological resources is designed too strictly, the establishment of the system will make the research and development staffs and related institutions hang back with hesitation both domestically and internationally. Their intention of bioprospecting in the designated country will then be reduced. On the other hand, if the system is designed too loosely, it will not be able to protect the rights of the owner of the resources. As a result, currently, every country holds a cautious attitude in setting up the regulations of managing the access to biological resources. Currently, many countries and regional international organizations already set up ABS system, such as Andean Community, African Union, Association of Southeast Asia Nations (ASEAN), Australia, South Africa, and India, all are enthusiastic with the establishment of the regulations regarding the access management of biological resources and genetic resources. On the other hand, there are still many countries only use traditional and existing conservation-related regulations to manage the access of biological resources. Since it has been more than 10 years that the regulation of access to biological resources and benefit sharing has been developed in some countries, how is Taiwan's current situation regarding this issue? Taiwan's Existing Regulations on the Access to Bioloical Resources In terms of regulations, Taiwan's existing management style of the access to biological resources is similar to that of the US and the EU. It refers to the existing regulations on environmental protection and conservation, and evaluates from the perspective of environmental protection to control and manage the exploitation and application of the related biological resources. These regulations include the Wildlife Conservation Act, theNational Park Law, the Forestry Act, the Cultural Heritage Preservation Act, and the Aboriginal Basic Act. The paragraphs below describe the contents of the acts mentioned that are related to the access to biological resources. 1 、 Wildlife Conservation Act According to the Wildlife Conservation Act, the Protected Species and the products made of cannot be hunted, traded, owned, imported, exported, raised, bred, and processed unless the number of protected wildlife has exceeded the amount the environment permits, or carry the objectives of academic research and education with the permits of central or regional authorities. As for the hunting of General Wildlife, pre-application and approval is needed with the exception of projects based on the objectives of academic research or education. In addition, the import and export of the living wildlife and the products of Protected Wildlife are restricted to the condition of being permitted by management authority. With respect to the import and export of living Protected Wildlife, Academic research institutions and colleagues are the only person who can seek for the approval of management authority before they proceed. 2 、 National Park Law The design and management of Taiwan's national parks are based on the regulations listed in the National Park Law with the purpose of protecting our country's exclusive natural scenery, wildlife and historical spots. Based on the properties and the nature of resources, the national park management structures the preserved area into general control area, playground and resting area, spot of historical interest, special landmark area, and ecological protection area. Ecological protection area refers to the areas where the natural surroundings, creatures, the society they live and propagate are strictly protected only for the research of ecology. According to the regulation of National Park Law, inside the national park area, it is prohibited to hunt animals, fish, take off flowers or trees, not to mention the behaviours that are prohibited by the management authority. Exceptions are made based on the conditions of preserved areas and for the research purposes. In the general control area or resting area, the national park authority allows fishing or other activities agreed by the authority. However, these activities are prohibited in the preserved area of historical interest, special landmark area, and ecological protection area. To suit special purposes, in the special landmark area or ecological protection area, collection of specimens is allowed subject to the approval of authority. Under the purpose of academic research, better management of public safety, and special management of national park, the Ministry of the Interior will permit the collection of specimen. However, to enter the ecological protection area, one must obtain the permission of authority. 3 、 Forestry Act To protect the forest resources and to maximize the public welfare and economic effectiveness, some of the properties are classified as forestry land and being managed by authority. Based on the Forestry Act, management authority has to restrict the area of cutting timber and to identify the area or period of restricted digging of greensward, tree roots, and grass roots, based on the condition of the forest. In addition, to maintain the current ecological environment in the forest, and to preserve the diversity of species, identification of natural preserved area is needed inside the forestland. The entrance and exit of human and vehicles are controlled based on the nature of the resources inside the preserve area. Unless obtaining the approval from the management authority, not a single activity of damaging, logging or digging soil, stones, greensward and roots is allowed. Furthermore, any unauthorized activity of collecting specimen inside the forest recreation area and natural reserve will be fined. Collecting flowers and plants in these areas, or trespassing the natural reserve will also be fined. Only the activities taken by the aborigines to sustain their living or accommodate their customs are not restricted. 4 、 Cultural Heritage Preservation Act The objectives of setting up the Cultural Heritage Preservation Act are to preserve and apply the cultural resources, to enrich the spiritual lives of citizens, and to add glory to the existing diverse cultures. The Cultural Heritage Preservation Act classifies the natural landscape and scenery as cultural assets. Vistas of Natural Culture refer to the natural areas, landforms, plants and mineral which contain the values of preservation. It can be further grouped into natural preserved area and natural monuments. Since the natural monuments include the unusual plants and mineral, it is connected to the management of biological resources. According to the Cultural Heritage Preservation Act, unless approved by the management authority, it is prohibited to collect, log, destroy the plants or bio resources classified as natural monuments or trespass into the area of natural preserve. For the purposes of academic research, or for the memorial ceremony of aboriginal custom, research institute and the aborigines can collect the natural monuments without the approval of authority. 5 、 Aboriginal Basic Act To protect the basic rights of the aborigines, and to sustain and develop the aboriginal society, the Aboriginal Basic Act was designed and enacted. The government not only admits the aborigine's rights in lands and natural resources, but also permits the non-profit behaviour such as hunting of wildlife, colleting of wild plants and fungi for the objective of complying with traditional culture, ceremony or private uses. In addition, the Aboriginal Basic Act provides the requirement of Prior Informed Consent (PIC) to require government or private individual to inform the aborigines before they proceed with land development, resource exploitation, ecological preservation, and academic research in the land where the aborigines live. They need to consult and obtain the aborigines' agreement or participation, and to share the related interests derived from this project. In the case of restricting the aborigine's right of the use of land or natural resources by law, the government,shall consult with the aborigines or the tribe and reach the agreement. When the government wish to design and establish national park, national scenic area, forestry area, ecological protection area, recreational area, or other resource management authorities, the government should obtain the agreement from the local aborigines and to build up the co-management mechanism.
Impact of Government Organizational Reform to Research Legal System and Response Thereto (2) – Observation of the Swiss Research Innovation SystemImpact of Government Organizational Reform to Research Legal System and Response Thereto (2) – Observation of the Swiss Research Innovation System I. Foreword Switzerland is a landlocked country situated in Central Europe, spanning an area of 41,000 km2, where the Alps occupy 60% of the territory, while it owns little cultivated land and poor natural resources. In 2011, its population was about 7,950,000 persons[1]. Since the Swiss Federal was founded, it has been adhering to a diplomatic policy claiming neutrality and peace, and therefore, it is one of the safest and most stable countries in the world. Switzerland is famous for its high-quality education and high-level technological development and is very competitive in biomedicine, chemical engineering, electronics and metal industries in the international market. As a small country with poor resources, the Swiss have learnt to drive their economic and social development through education, R&D and innovation a very long time ago. Some renowned enterprises, including Nestle, Novartis and Roche, are all based in Switzerland. Meanwhile, a lot of creative small-sized and medium-sized enterprises based in Switzerland are dedicated to supporting the export-orientation economy in Switzerland. Switzerland has the strongest economic strength and plentiful innovation energy. Its patent applications, publication of essay, frequencies of quotation and private enterprises’ innovation performance are remarkable all over the world. According to the Global Competitiveness Report released by the World Economic Forum (WEF), Switzerland has ranked first among the most competitive countries in the world for four years consecutively since 2009[2]. Meanwhile, according to the Global Innovation Index (GII) released by INSEAD and the World Intellectual Property Organization (WIPO) jointly, Switzerland has also ranked first in 2011 and 2012 consecutively[3]. Obviously, Switzerland has led the other countries in the world in innovation development and economic strength. Therefore, when studying the R&D incentives and boosting the industrial innovation, we might benefit from the experience of Switzerland to help boost the relevant mechanism in Taiwan. Taiwan’s government organization reform has been launched officially and boosted step by step since 2012. In the future, the National Science Council will be reformed into the “Ministry of Science and Technology”, and the Ministry of Economic Affairs into the “Ministry of Economy and Energy”, and the Department of Industrial Development into the “Department of Industry and Technology”. Therefore, Taiwan’s technology administrative system will be changed materially. Under the new government organizational framework, how Taiwan’s technology R&D and industrial innovation system divide work and coordinate operations to boost the continuous economic growth in Taiwan will be the first priority without doubt. Support of innovation policies is critical to promotion of continuous economic growth. The Swiss Government supports technological research and innovation via various organizations and institutions effectively. In recent years, it has achieved outstanding performance in economy, education and innovation. Therefore, we herein study the functions and orientation of the competent authorities dedicated to boosting research and innovation in Switzerland, and observe its policies and legal system applied to boost the national R&D in order to provide the reference for the functions and orientation of the competent authorities dedicated to boosting R&D and industrial innovation in Taiwan. II. Overview of Swiss Federal Technology Laws and Technology Administrative System Swiss national administrative organization is subject to the council system. The Swiss Federal Council is the national supreme administrative authority, consisting of 7 members elected from the Federal Assembly and dedicated to governing a Federal Government department respectively. Switzerland is a federal country consisting of various cantons that have their own constitutions, councils and governments, respectively, entitled to a high degree of independence. Article 64 of the Swiss Federal Constitution[4] requires that the federal government support research and innovation. The “Research and Innovation Promotion Act” (RIPA)[5] is dedicated to fulfilling the requirements provided in Article 64 of the Constitution. Article 1 of the RIPA[6] expressly states that the Act is enacted for the following three purposes: 1. Promoting the scientific research and science-based innovation and supporting evaluation, promotion and utilization of research results; 2. Overseeing the cooperation between research institutions, and intervening when necessary; 3. Ensuring that the government funding in research and innovation is utilized effectively. Article 4 of the RIPA provides that the Act shall apply to the research institutions dedicated to innovation R&D and higher education institutions which accept the government funding, and may serve to be the merit for establishment of various institutions dedicated to boosting scientific research, e.g., the National Science Foundation and Commission of Technology & Innovation (CTI). Meanwhile, the Act also provides detailed requirements about the method, mode and restriction of the government funding. According to the RIPA amended in 2011, the Swiss Federal Government’s responsibility for promoting innovation policies has been extended from “promotion of technology R&D” to “unification of education, research and innovation management”, making the Swiss national industrial innovation framework more well-founded and consistent[8] . Therefore, upon the government organization reform of Switzerland in 2013, most of the competent authorities dedicated to technology in Swiss have been consolidated into the Federal Department of Economic Affairs, Education and Research. Under the framework, the Swiss Federal Government assigned higher education, job training, basic scientific research and innovation to the State Secretariat for Education, Research and Innovation (SERI), while the Commission of Technology & Innovation (CTI) was responsible for boosting the R&D of application scientific technology and industrial technology and cooperation between the industries and academy. The two authorities are directly subordinate to the Federal Department of Economic Affairs, Education and Research (EAER). The Swiss Science and Technology Council (SSTC), subordinate to the SERI is an advisory entity dedicated to Swiss technology policies and responsible for providing the Swiss Federal Government and canton governments with the advice and suggestion on scientific, education and technology innovation policies. The Swiss National Science Foundation (SNSF) is an entity dedicated to boosting the basic scientific R&D, known as the two major funding entities together with CTI for Swiss technology R&D. The organizations, duties, functions and operations of certain important entities in the Swiss innovation system are introduced as following. Date source: Swiss Federal Department of Economic Affairs, Education and Research official website Fig. 1 Swiss Innovation Framework Dedicated to Boosting Industries-Swiss Federal Economic, Education and Research Organizational Chart 1. State Secretariat of Education, Research and Innovation (SERI) SERI is subordinate to the Department of Economic Affairs, Education and Research, and is a department of the Swiss Federal Government dedicated to managing research and innovation. Upon enforcement of the new governmental organization act as of January 1, 2013, SERI was established after the merger of the State Secretariat for Education and Research, initially subordinate to Ministry of Interior, and the Federal Office for Professional Education and Technology (OEPT), initially subordinated to Ministry of Economic Affairs. For the time being, it governs the education, research and innovation (ERI). The transformation not only integrated the management of Swiss innovation system but also unified the orientations toward which the research and innovation policy should be boosted. SERI’s core missions include “enactment of national technology policies”, “coordination of research activities conducted by higher education institutions, ETH, and other entities of the Federal Government in charge of various areas as energy, environment, traffic and health, and integration of research activities conducted by various government entities and allocation of education, research and innovation resources. Its functions also extend to funding the Swiss National Science Foundation (SNSF) to enable SNSF to subsidize the basic scientific research. Meanwhile, the international cooperation projects for promotion of or participation in research & innovation activities are also handled by SERI to ensure that Switzerland maintains its innovation strength in Europe and the world. The Swiss Science and Technology Council (SSTC) is subordinate to SERI, and also the advisory unit dedicated to Swiss technology policies, according to Article 5a of RIPA[9]. The SSTC is responsible for providing the Swiss Federal Government and canton governments with advice and suggestion about science, education and innovation policies. It consists of the members elected from the Swiss Federal Council, and a chairman is elected among the members. 2. Swiss National Science Foundation (SNSF) The Swiss National Science Foundation (SNSF) is one of the most important institutions dedicated to funding research, responsible for promoting the academic research related to basic science. It supports about 8,500 scientists each year. Its core missions cover funding as incentives for basic scientific research. It grants more than CHF70 million each year. Nevertheless, the application science R&D, in principle, does not fall in the scope of funding by the SNSF. The Foundation allocates the public research fund under the competitive funding system and thereby maintains its irreplaceable identity, contributing to continuous output of high quality in Switzerland. With the support from the Swiss Federal Government, the SNSF was established in 1952. In order to ensure independence of research, it was planned as a private institution when it was established[10]. Though the funding is provided by SERI, the SNSF still has a high degree of independence when performing its functions. The R&D funding granted by the SNSF may be categorized into the funding to free basic research, specific theme-oriented research, and international cooperative technology R&D, and the free basic research is granted the largest funding. The SNSF consists of Foundation Council, National Research Council and Research Commission[11]. Data source: prepared by the Study Fig. 2 Swiss National Science Foundation Organizational Chart (1) Foundation Council The Foundation Council is the supreme body of the SNSF[12], which is primarily responsible for making important decisions, deciding the role to be played by the SNSF in the Swiss research system, and ensuring SNSF’s compliance with the purpose for which it was founded. The Foundation Council consists of the members elected from the representatives from important research institutions, universities and industries in Swiss, as well as the government representatives nominated by the Swiss Federal Council. According to the articles of association of the SNSF[13], each member’s term of office should be 4 years, and the members shall be no more than 50 persons. The Foundation Council also governs the Executive Committee of the Foundation Council consisting of 15 Foundation members. The Committee carries out the mission including selection of National Research Council members and review of the Foundation budget. (2) National Research Council The National Research Council is responsible for reviewing the applications for funding and deciding whether the funding should be granted. It consists of no more than 100 members, mostly researchers in universities and categorized, in four groups by major[14], namely, 1. Humanities and Social Sciences; 2. Math, Natural Science and Engineering; 3. Biology and Medical Science; and 4. National Research Programs (NRPs)and National Centers of Competence in Research (NCCRs). The NRPs and NCCRs are both limited to specific theme-oriented research plans. The funding will continue for 4~5years, amounting to CHF5 million~CHF20 million[15]. The specific theme-oriented research is applicable to non-academic entities, aiming at knowledge and technology transfer, and promotion and application of research results. The four groups evaluate and review the applications and authorize the funding amount. Meanwhile, the representative members from each group form the Presiding Board dedicated to supervising and coordinating the operations of the National Research Council, and advising the Foundation Council about scientific policies, reviewing defined funding policies, funding model and funding plan, and allocating funding by major. (3) Research Commissions Research Commissions are established in various higher education research institutions. They serve as the contact bridge between higher education academic institutions and the SNSF. The research commission of a university is responsible for evaluating the application submitted by any researcher in the university in terms of the school conditions, e.g., the school’s basic research facilities and human resource policies, and providing advice in the process of application. Meanwhile, in order to encourage young scholars to attend research activities, the research committee may grant scholarships to PhD students and post-doctor research[16]. ~to be continued~ [1] SWISS FEDERAL STATISTICS OFFICE, Switzerland's population 2011 (2012), http://www.bfs.admin.ch/bfs/portal/en/index/news/publikationen.Document.163772.pdf (last visited Jun. 1, 2013). [2] WORLD ECONOMIC FORUM [WEF], The Global Competiveness Report 2012-2013 (2012), http://www3.weforum.org/docs/WEF_GlobalCompetitivenessReport_2012-13.pdf (last visited Jun. 1, 2013); WEF, The Global Competiveness Report 2011-2012 (2011), http://www3.weforum.org/docs/WEF_GCR_Report_2011-12.pdf (last visited Jun. 1, 2013); WEF, The Global Competiveness Report 2010-2011 (2010), http://www3.weforum.org/docs/WEF_GlobalCompetitivenessReport_2010-11.pdf (last visited Jun. 1, 2013); WEF, The Global Competiveness Report 2009-2010 (2009),. http://www3.weforum.org/docs/WEF_GlobalCompetitivenessReport_2009-10.pdf (last visited Jun. 1, 2013). [3] INSEAD, The Global Innovation Index 2012 Report (2012), http://www.globalinnovationindex.org/gii/GII%202012%20Report.pdf (last visited Jun. 1, 2013); INSEAD, The Global Innovation Index 2011 Report (2011), http://www.wipo.int/freepublications/en/economics/gii/gii_2011.pdf (last visited Jun. 1, 2013). [4] SR 101 Art. 64: “Der Bund fördert die wissenschaftliche Forschung und die Innovation.” [5] Forschungs- und Innovationsförderungsgesetz, vom 7. Oktober 1983 (Stand am 1. Januar 2013). For the full text, please see www.admin.ch/ch/d/sr/4/420.1.de.pdf (last visited Jun. 3, 2013). [6] Id. [7] Id. [8] CTI, CTI Multi-year Program 2013-2016 7(2012), available at http://www.kti.admin.ch/?lang=en&download=NHzLpZeg7t,lnp6I0NTU042l2Z6ln1ad1IZn4Z2qZpnO2Yuq2Z6gpJCDeYR,hGym162epYbg2c_JjKbNoKSn6A-- (last visited Jun. 3, 2013). [9] Supra note 5. [10] Swiss National Science Foundation, http://www.snf.ch/E/about-us/organisation/Pages/default.aspx (last visited Jun. 3, 2013). [11] Id. [12] Foundation Council, Swiss National Science Foundation, http://www.snf.ch/E/about-us/organisation/Pages/foundationcouncil.aspx (last visited Jun. 3, 2013). [13] See Statutes of Swiss National Science Foundation Art.8 & Art. 9, available at http://www.snf.ch/SiteCollectionDocuments/statuten_08_e.pdf (last visited Jun. 3, 2013). [14] National Research Council, Swiss National Science Foundation, http://www.snf.ch/E/about-us/organisation/researchcouncil/Pages/default.aspx (last visted Jun.3, 2013). [15] Theres Paulsen, VISION RD4SD Country Case Study Switzerland (2011), http://www.visionrd4sd.eu/documents/doc_download/109-case-study-switzerland (last visited Jun.6, 2013). [16] Research Commissions, Swiss National Science Foundation, http://www.snf.ch/E/about-us/organisation/Pages/researchcommissions.aspx (last visted Jun. 6, 2013).