|
THE NATIONAL ASSEMBLY |
THE SOCIALIST REPUBLIC OF VIETNAM |
|
Law No.: 134/2025/QH15 |
Hanoi, December 10, 2025 |
Pursuant to the Constitution of the Socialist Republic of Vietnam amended by the Resolution No. 203/2025/QH15;
The National Assembly of Vietnam promulgates the Law on Artificial Intelligence.
1. This Law regulates the research, development, provision, deployment and use of the system of artificial intelligence (hereinafter referred to as “AI activities”); rights and obligations of relevant organizations and individuals; and state management of AI activities in Vietnam.
2. AI activities only serving national defense, security and cipher activities are not under the scope of this Law.
This Law applies to authorities, organizations and individuals in Vietnam; and foreign organizations and individuals participating in AI activities in Vietnam.
For the purposes of this Law, the terms below shall be construed as follows:
1. Artificial intelligence (hereinafter referred to “AI”) is the electronic implementation of human intellectual capacities including learning, reasoning, perception, judgment and understanding natural language.
2. An AI system is a machine-based system designed to perform AI capabilities with different levels of autonomy that is capable of self-adaptation after deployment. Based on clearly defined or implicitly formed objectives, the system infers from input data to generate outputs such as predictions, content, recommendations or decisions that may affect the physical or digital environment.
3. A developer is an organization or individual engaging in designing, developing, training, testing or adjusting all or part of a model, algorithm or system of AI; and having the direct control over the technical methods, training data or model parameters.
4. A provider is an organization and individual who provides an AI system to the market, or puts it into operation under their own name, brand or trademark regardless of whether the system is developed by themselves or by a third party.
5. A deployer is an organization and individual using an AI system under their control in professional, commercial and service activities, excluding the use for personal or non-commercial purposes.
6. A user is an organization and individual who directly interact with the AI system or use its output.
7. An affected person is an organization and individual that is directly or indirectly affected to their legitimate rights, interests, life, health, property, reputation or opportunities to access to services due to the deployment or output of the AI system.
8. A serious incident is an event occurring during the operation of an AI system that causes or poses a risk of significant harm to life, health, human rights, property, cybersecurity, public order, environment, or disrupts the operation of important information systems related to national security.
Article 4. Primary principles for AI activities
1. Apply human-centered approach; protect human rights, privacy, national interests, public interests and national security; comply with the Constitutions and laws.
2. Serve the human; not replace human authority and responsibilities. Maintain human control and intervention to all decisions and behaviors of AI systems; to system security, data security and information confidentiality; to capabilities for inspection and supervision of the development and operation of AI systems.
3. Ensure the fairness, transparency, impartiality, non-discrimination, and cause no harm to the human or society; comply with Vietnamese ethical standards and cultural values; be accountable for decisions and consequences of AI systems.
4. Promote the development of green, inclusive and sustainable AI; encourage the development and application of AI in order to use energy efficiently, save resources and reduce negative environmental impacts.
Article 5. State policies for AI activities
1. Implement policies for development of AI as a key momentum for national growth, innovation and sustainable development.
2. Encourage regulatory sandboxes for technology; apply management measures in conformity with the level of risk; and encourage voluntary compliance mechanisms.
3. Implement policies to ensure rights, and facilitate access, learning and benefits from AI for organizations and individuals; encourage the development and application of AI to serve social welfare, support people with disabilities, the poor and ethnic minorities to narrow the digital divide; preserve, promote and maintain the national cultural identity.
4. Prioritize investment and mobilize social resources to develop data infrastructure, computing infrastructure, safe AI, high-quality human resources and shared AI platforms of national strategic importance.
5. Prioritize the application of AI in management, administration, public services and decision-making of state authorities to enhance efficiency, transparency and service quality for citizens and enterprises; encourage application of AI across socio-economical sectors to improve productivity, service quality and management effectiveness.
6. Promote the organization, networks and social initiatives to promote safety, ethics, trustworthiness, and build social trust in AI development.
7. Promote the application of AI in business activities and key socio-economical sectors; develop startup and innovation ecosystems; encourage public-private partnerships.
8. Proactively integrate and cooperate internationally; participate in developing and shaping global standards and administration frameworks; ensure national interests and sovereignty regarding AI.
Article 6. Application of AI in sectors
1. The application of AI in sectors shall comply with principles for management of risks in accordance with this Law and relevant laws.
2. For essential sectors that directly impact human life, health, legitimate rights and interests, or social order and safety, the application of AI shall be managed with stricter risk control in accordance with characteristics of each sector including:
a) Health: ensure patient safety and reliability under practical use conditions; protect data on health as prescribed by laws;
b) Education: ensure suitability with the age characteristics and development of learners; prevent risks in evaluation, classification and impacts on learners; ensure data security and privacy rights.
3. The application of AI in scientific research activities shall comply with research ethics, scientific integrity, and prevent fraud and plagiarism during research and publication of results.
4. The Government and ministerial authorities shall, within their functions, tasks and powers, elaborate requirements for safety, risk safety and conditions for deployment of AI in sectors under their management, and comply with this Law.
Article 7. Prohibited activities
1. Exploit or appropriate AI systems to commit illegal act, or infringe upon the legitimate rights and interests of organizations and individuals.
2. Develop, provide, deploy or use AI systems for the following purposes:
a) Committing acts prohibited by laws;
b) Using forged or simulated elements of real people or events to deliberately and systematically deceive or manipulate human perception and behaviors that cause serious harm to legitimate rights and interests of individuals;
c) Exploiting the vulnerabilities of vulnerable groups including children, the elderly, persons with disabilities, ethnic minorities or persons lacking legal capacity, persons with limited legal capacity, or persons with cognitive or behavioral control difficulties to cause harm to themselves or others;
d) Creating or disseminating fabricated content that may cause serious damages to national security, social order and safety.
3. Collect, handle or use data to develop, train, testing or operate AI systems against the laws on data, protection of personal data, intellectual property and cybersecurity.
4. Obstruct, disable or falsify supervision, intervention and control mechanisms of the human over AI systems as prescribed by laws.
5. Conceal information that shall be disclosed, transparent or explained; erase or falsify mandatory information, labels or warnings in AI activities.
6. Exploit research, sandbox, evaluation or inspection activities of AI systems to commit acts against the laws.
Article 8. Single-window website on AI and National database on AI systems
1. The single-window website on AI is a digital platform established to support the reception and registration for regulatory sandboxes; receive notifications of AI system classification results, serious incident reports, and periodic reports; disclose information about AI systems, conformity evaluation results, violation handling results as prescribed by laws; and connect support programs, funds, infrastructure and shared data.
2. The National database on AI systems shall be developed and managed to serve management, supervision and disclosure of information about AI systems in accordance with the laws.
3. The disclosure, connection and share of data on the single-window website on AI and the National database on AI systems shall comply with regulations on information security; protection of state secrets, business secrets and personal data.
4. The Government shall elaborate mechanisms for operation, management and use of the single-window website on AI and the National database on AI systems.
RISK-BASED CLASSIFICATION AND MANAGEMENT OF AI SYSTEMS
Article 9. Classification of risk levels of AI systems
1. AI systems shall be classified according to the following risk levels:
a) High-risk AI systems are those causing significant harm to life, health, legitimate rights and interests of organizations and individuals, national interests, public interests or national security;
b) Medium-risk AI systems are those causing confusion, influence or manipulate users due to the inability to recognize that the interacting entity is an AI system or the content generated by the system;
c) Low-risk AI systems are those that are not prescribed in points a and b of this clause.
2. The classification of risk of AI systems shall be determined based on criteria regarding impacts on human rights, safety and security; the sector of application, especially essential sectors or those related to public interests; and the scope of users and the scale of the system’s influence.
3. The Government shall elaborate this Article.
Article 10. Classification and notification on AI systems
1. Providers shall classify AI systems before putting them into operation. Systems classified as medium-risk or high-risk shall be accompanied with classification dossiers.
2. Deployers may inherit classification results of providers and are responsible for protecting the safety and integrity of the systems during use. In cases of adjustments, integration or changes to functions that generate new or higher risks, they shall cooperate with providers in reclassifying the systems.
3. For medium-risk or high-risk systems, providers shall notify the classification results to the Ministry of Science and Technology via the single-window website on AI before putting them into operation. Organizations and individuals developing low-risk AI systems are encouraged to disclose basic system information to increase the transparency.
4. If the risk level cannot be determined, providers may request guidelines for classification from the Ministry of Science and Technology based on technical dossiers.
5. The inspection and supervision shall be conducted according to the risk level of the systems:
a) High-risk AI systems shall be inspected periodically or when there are signs of violations;
b) Medium-risk AI systems shall be supervised via reports, sample inspections, or evaluations by independent organizations;
c) Low-risk AI systems shall be supervised and inspected upon incidents, complaints, or when necessary to ensure safety without causing unnecessary obligations for organizations or individuals.
6. Based on inspection and supervision results prescribed in clause 5 of this Article, when discrepancies or dishonest declarations are discovered, competent authorities shall request re-classification, dossier supplementation, or suspension of use, and handle cases in accordance with the laws.
7. The Government shall elaborate the notification, procedures for notification and technical guidelines on risk classification.
Article 11. Transparency responsibility
1. Providers shall ensure that AI systems interacting directly with humans are designed and operated so that users can recognize when they are interacting with the systems, unless prescribed otherwise.
2. Providers shall ensure that audio, images and videos generated by AI systems are marked in a machine-readable format as prescribed by the Government.
3. Deployers are responsible for making notifications when texts, audio, images or videos created or edited by AI systems are provided if such content may cause confusion regarding the authenticity of events or characters, unless prescribed otherwise.
4. Deployers are responsible for ensuring that audio, images or videos created or edited by AI systems to simulate or imitate the appearance, voice of real persons, or reenact actual events are labeled clearly to distinguish them from human-made content.
For products that are cinematographic, artistic or creative works, the labeling prescribed in this clause shall be done in a manner that does not hinder the display, presentation or enjoyment of the works.
5. Providers and deployers shall maintain transparency information as prescribed in this Article throughout the process of providing the systems, products or content to users.
6. The Government shall elaborate methods for notification and labeling.
Article 12. Responsibilities for management and handling of AI incidents
1. Developers, providers, deployers and users of AI systems are responsible for ensuring the security, reliability; promptly detecting and fixing incidents that may cause harm to people, property or social order.
2. When a serious incident occurs, developers, providers, deployers and users have the following responsibilities:
a) Developers and providers shall promptly apply technical measures to fix, suspend or recall the system, and notify competent authorities;
b) Deployers and users shall promptly record, notify incidents and cooperate in the fixing process.
3. State management authorities shall receive, verify and provide guidelines for handling incidents. When necessary, they may suspend, recall or re-evaluate AI systems.
4. Reports and handling of incidents shall be carried out via the single-window website on AI.
5. The Government shall elaborate the reporting and responsibilities of relevant authorities, organizations and individuals in accordance with the severity of the incident and the impact of AI systems.
Article 13. Evaluation of the conformity of high-risk AI systems
1. The conformity of high-risk AI systems shall be evaluated in accordance with this Law before being put into operation or when there is a significant change during operation. If there are standards or technical regulations applicable to AI systems, the conformity evaluation shall also comply with such standards or technical regulations.
2. The conformity evaluation is the confirmation that AI systems comply with Article 14 of this Law and shall be carried out as follows:
a) For high-risk AI systems within the List of AI systems required conformity evaluation before being put into operation, the evaluation shall be carried out by conformity evaluation organizations registered or recognized in accordance with the laws;
b) For other high-risk AI systems, providers shall self-evaluate the conformity or hire conformity evaluation organizations registered or recognized in accordance with the laws.
3. The result of conformity evaluation is a requirement for a high-risk AI system to be permitted for operation. Organizations and individuals with systems that have undergone conformity evaluation are responsible for maintaining the conformity and disclosing the information as prescribed by the Government. The result is also a ground for inspection and supervision of the compliance with regulations in Article 10 of this Law.
4. The Government shall regulate the List of high-risk AI systems including the List of AI systems required conformity evaluation before being put into operation.
5. The evaluation of conformity and inspection of AI systems shall ensure the independence and sufficient technical competence as prescribed by laws, and be subject to periodic supervision of competent state authorities.
6. The Government shall elaborate this Article.
Article 14. Management of high-risk AI systems
1. Providers of high-risk AI systems are responsible for:
a) Promulgating and maintaining risk management measures, and regularly reviewing the systems when there is a significant change or a new risk arise;
b) Administer the training, testing and operational data to ensure the quality in accordance with technical capabilities and the system’s purpose;
c) Prepare, update and archive technical dossiers and activity logs at a level that is necessary for conformity evaluation and post-deployment inspection; provide this information to competent state authorities in accordance with necessary principles and inspection purpose without disclosing trade secrets;
d) Design the systems that ensure human oversight and intervention capabilities;
dd) Fulfill transparency obligations and perform incident handling as prescribed in Articles 11 and 12 of this Law;
e) Provide explanation to competent state authorities on the use purpose, operational principles regarding functional description, main types of input data, risk management and control measures, and other necessary information for inspection; provide users and affected persons with public information regarding functional description, operational methods and risk warnings during operation. The explanation and provision of information shall not include the disclosure of source codes, detailed algorithms, parameter sets or information related to trade secrets or technical secrets;
g) Cooperate with competent state authorities and deployers in inspecting, evaluating, post-inspecting and fixing incidents related to the systems.
2. Deployers of high-risk AI systems are responsible for:
a) Operating and supervising the systems in accordance with classified purpose, scope and risk level without generating new or higher risks;
b) Ensure data security and human intervention during operation;
c) Maintain the compliance with standards and technical regulations on AI during operation;
d) Fulfill transparency obligations and perform incident handling as prescribed in Articles 11 and 12 of this Law;
dd) Provide explanation to competent state authorities on operation, risk controlling measures, incident handling and necessary information inspection; provide users and affected persons with public information regarding functional description, operational methods and risk warnings during operation;
e) Cooperate with providers and competent state authorities in inspecting, evaluating, post-evaluating and fixing incidents.
3. Users of high-risk AI systems are responsible for complying operational procedures, technical guidelines and risk measures to ensure the safety; not illegally interfering to adjust features of systems; and promptly notifying incidents to deployers.
4. The explanation shall comply with technical capabilities of systems without disclosing trade secrets in accordance with the laws.
5. Providers and deployers are encouraged to participate in civil liability insurance or apply measures to fulfill other obligations to promptly fix incidents and pay damages.
6. Foreign providers having high-risk AI systems that are provided in Vietnam shall have a legal contact point in Vietnam. In cases where a system is subject to mandatory conformity certification before being put into operation, it must have a commercial presence or an authorized representative in Vietnam.
7. The Government shall elaborate this Article.
Article 15. Management of medium-risk and low-risk AI systems
1. Medium-risk AI systems shall be managed as follows:
a) Providers and deployers shall ensure the transparency in accordance with Article 11 of this Law;
b) Providers shall explain the use purpose, operational principles regarding functional description, main input data, risk management and safety measures of the systems at the request of state authorities during inspection or when there is a sign of risk or incident. The explanation shall not include the disclosure of source codes, detailed algorithms, parameter sets, trade secrets or technical secrets;
c) Deployers shall explain the operation, risk controlling, incident handling, protection of rights and legitimate interests of organizations and individuals at the request of competent state authorities during inspection or incident handling;
d) Users are responsible for complying with regulations on notification and labeling AI systems.
2. Low-risk AI systems shall be managed as follows:
a) Providers shall provide explanation at the request of competent state authorities when there is a sign of violation or such systems affect rights and legitimate interests of organizations and individuals;
b) Deployers shall provide explanation at the request of competent state authorities when there is a sign of violation or such systems affect rights and legitimate interests of organizations and individuals;
c) Users may use systems for legal purposes and take responsibilities for their use.
3. The State shall encourage organizations and individuals to deploy medium-risk and low-risk AI systems in accordance with technical regulations on AI.
DEVELOPMENT OF INFRASTRUCTURE AND ASSURANCE OF NATIONAL AI SOVEREIGNTY
Article 16. Infrastructure of national AI
1. The national AI infrastructure is a strategic infrastructure including infrastructure invested by the State, enterprises and social organizations. It is developed as a uniform, open and secure ecosystem that is capable of connectivity, share and expansion in accordance with requirements for development and application of AI.
2. The State shall provide guidelines, coordinate and ensure infrastructure capacity to serve national AI development; encourage enterprises, research institutes, universities and social organizations to invest, develop and share infrastructure; promote public-private partnerships in developing AI infrastructure.
3. The State shall invest, develop and operate AI infrastructure for provision of public services, support for research, development, state management and startup including shared computing capacity and data; platforms for training, testing and sandbox; platform models, multipurpose AI models, large language models in Vietnamese and ethnic minority languages; other infrastructure components.
4. The national AI infrastructure invested by the State, enterprises and social organizations shall be connected, shared and used in accordance with standards, technical regulations and requirements regarding safety, security and protection of data.
5. The application of important AI in essential sectors in accordance with the list issued by the Prime Minister shall be deployed by national AI infrastructure to ensure security and controllability.
6. The Government shall elaborate the mechanisms for coordination, share, incentive and measures for development of national AI infrastructure for each stage in accordance with requirements for national security.
1. The Database for AI is an important component of the national AI infrastructure including the National Database; databases of ministries, ministerial authorities, governmental authorities, People’s Committees at all levels; databases of organizations and individuals. The Database for AI is developed, managed and used for training, testing, evaluating and developing AI in accordance with laws on data, protection of personal data and intellectual property.
2. The National Database for AI shall be invested by the State, developed and operated at the National Data Center; organized in accordance with principles of openness, safety and control, requirements for quality, connectivity and use. It includes open data, conditional access data and commercial data as prescribed by laws.
3. The Database for AI of ministries, ministerial authorities, governmental authorities and People’s Committees at all levels shall be developed, updated and connected to the National Database for AI in accordance with standards, technical regulations, data quality and information security.
4. Organizations and individuals are encouraged to share their databases for AI with state authorities, other organizations and individuals upon reaching agreement in accordance with laws on data, personal data protection, intellectual property, and protection of rights and legitimate interests of relevant parties.
5. The Prime Minister shall promulgate the List of dataset for development of AI in essential sectors prioritizing data on culture, Vietnamese language, ethnic minority languages, administrative procedures, health, education, agriculture, environment, transportation, socio-economic affairs and other important sectors.
6. The Government shall elaborate principles for connecting, mechanisms for sharing, using and protecting data in the Database for AI.
Article 18. Mastery over AI technology
1. The State shall prioritize the development and mastery of core AI technologies; prioritize resources for researching and developing multipurpose AI models, large language models in Vietnamese and ethnic minority languages, Vietnamese knowledge processing technologies, high-performance computing and training technologies, hardware, and semiconductors to serve AI; promote the development and application of open-source software to enhance technological autonomy, national safety and sovereignty in the digital environment.
2. The State shall promote research, development, completion and application of domestic AI; support organizations and individuals develop models, algorithms, software, hardware and platform technology; encourage measures that are resource-efficient and easily deployable in accordance with Vietnam’s conditions; develop national endogenous capacity and an innovation system for AI; and promote public-private partnerships to master the technology.
3. Organizations and individuals researching, developing and mastering core AI technologies are eligible for incentives and special mechanisms as prescribed by laws.
4. The State shall promote the application of AI for scientific research, analysis and simulation, technological design and sandbox, automation of researching procedures, development and innovation to improve national scientific and technological capacity; and facilitate the development of innovation and mastery over the lifecycle of AI technology.
5. The Government shall elaborate the mechanisms, criteria and methods for mastery over AI technology for each stage in accordance with requirements for national security.
APPLICATION OF AI, DEVELOPMENT OF INNOVATION ECOSYSTEM AND HUMAN RESOURCES
Article 19. National strategy for AI
1. The Prime Minister shall promulgate the National strategy for AI (hereinafter referred to as “Strategy”), review, evaluate and update at least once per 03 years, or when the technology or market has a significant change. Ministries, ministerial authorities, Governmental authorities and People’s Committees at all levels are responsible for integrating targets and tasks of the Strategy into strategies and plans for development of sectors, provinces/communes, and ensuring human resources for implementation.
2. The Strategy shall be developed based on development orientations for technology, infrastructure, data and human resources; promotion of research, mastery and application of AI in priority sectors; safety, innovation and national sovereignty in the digital environment. The Strategy shall regulate a system of indicators, methods and mechanisms for evaluation of the level of development of national AI.
3. The State shall encourage the development of groups of AI technology in accordance with Vietnam’s conditions that have the potential to create added value, are environmentally friendly, easy to widely apply, and contribute to the protection of national sovereignty in the digital environment.
Article 20. Development of AI ecosystem and market
1. Organizations and individuals engaging in AI activities are eligible for highest incentives and support in accordance with laws on science and technology, investment, digital technology industry, high technology and relevant laws; support for access to infrastructure, data and sandbox for research, production and commercialization of AI products and services.
2. The State shall support the development of AI ecosystem and market including:
a) Prioritize the use of AI products and services in accordance with laws on bidding;
b) Develop the market of AI products and services including technology exchange platforms and supply-demand connection platforms;
c) Ensure fair and transparent access to computing infrastructure, data and sandbox;
d) Apply incentives regarding tax, investment and finance based on the principles of encouraging research, production and commercialization of AI products and services.
3. The State shall encourage the development and application of new generation AI; promote the innovation and improve capabilities regarding administration, production, business and public services.
4. Organizations, individuals, enterprises, research facilities and state authorities are encouraged to use, share and recycle the data in the National database for AI to serve the research, training, testing and innovation in accordance with the laws on data, personal data protection, cybersecurity and intellectual property.
5. Small and medium enterprises, startups engaging in AI activities are prioritized to access the infrastructure, data and sandbox; eligible for support for fees, training and connecting to market for development of AI products and services.
6. The Government shall elaborate mechanisms, requirements and procedures for carrying out measures for development of AI ecosystem and market.
Article 21. AI regulatory sandbox
1. AI regulatory sandbox shall comply with the laws on science, technology and innovation, and regulations prescribed in clauses 2, 3 and 4 of this Article.
2. The result from the sandbox is the ground for competent state authorities to consider:
a) For recognition of the conformity evaluation result as prescribed in this Law;
b) For exemption, reduction or adjustment of obligations as prescribed in this Law.
3. Competent state authorities shall take charge and cooperate with relevant authorities in receiving, appraising and processing dossiers in accordance with fast appraisal and response procedures; supervise the sandbox, and issue the decision on suspension or termination of the sandbox when there is a risk to the security, rights or legitimate interests of organizations and individuals.
4. The Government shall elaborate this Article.
Article 22. Fund for development of national AI
1. The Fund for development of national AI (hereinafter referred to as “Fund”) is a non-budgetary state financial fund operating on a non-profit basis that is established by the Government to mobilize, coordinate and allocate resources to promote research, development, application and management of AI for socio-economic development, national defense, security, and enhancing national competitiveness.
2. The Fund shall be funded by the state budget; contributions, aids and sponsorships of domestic and foreign organizations and individuals; and other legal sources as prescribed by laws.
3. The Fund may apply a special financial mechanism accepting risks in science, technology, and innovation; flexibly allocate the capital according to the progress and requirements of the implementation without considering the fiscal year; apply simplified procedures for strategic tasks or fast deployment. The Fund is prioritized for investment, sponsorship and support in:
a) Development of AI infrastructure;
b) Research, development and mastery over core AI technologies;
c) Development of AI enterprises;
d) Training and attracting AI human resources;
dd) Other investment and support tasks for objectives of AI development as prescribed by the Government.
4. The Fund shall operate based on principles of openness, transparency, efficiency and proper purposes; ensure coordination and avoid overlap with other state financial funds.
5. The Government shall elaborate special financial mechanisms, and mechanisms for establishment, management and supervision of the Fund.
Article 23. Development of AI human resources
1. The State shall develop AI human resources comprehensively and in an interconnected manner across educational levels and training qualifications to establish a high-quality workforce to serve research, development, application and management of AI.
2. General education shall integrate basic knowledge on AI, computational thinking, digital skills and technology ethics into the compulsory curriculum. Activities for experience, research and creativity in the field of AI are encouraged.
3. Vocational education and higher education institutions are encouraged to develop training programs on AI, data science and relevant fields. Cooperation with enterprises, research institutes and international organizations in training, internship, research and technology transfer is also encouraged.
4. The State shall implement the National program for development of AI human resources including policies on training, scholarships, attracting and prioritizing experts; developing teams of lecturers, scientists and management personnel in the field of AI.
5. Organizations, training institutions, research institutes and enterprises engaging in developing AI human resources shall enjoy incentives and preferential mechanisms as prescribed by laws, while also being responsible for cooperating in training, applied research and professional practice, and linking education with practical needs.
6. Higher education institutions, research institutes and innovation centers are responsible for cooperating, sharing knowledge and participating in national and international networks for training, research and development of AI human resources.
7. The Ministry of Education and Training shall take charge in developing and proposing the Prime Minister for promulgation of the National program for development of AI human resources which regulates standards, recognition of training programs, mechanisms for mobilizing resources and incentive policies for organizations and individuals.
Article 24. Development of AI linkage cluster
1. The AI linkage cluster is a cooperative network between enterprises, research institutes, universities and relevant organizations organized to enhance linkage in functions, AI infrastructure and physical spaces to promote innovation, AI development, and improve competitiveness.
2. The State shall encourage the development of AI linkage clusters according to the model combining centralized physical spaces and digital linkage networks; establish cluster centers in high-tech zones, concentrated digital technology zones and innovation centers; attract organizations and individuals to invest in building technical infrastructure for cluster activities including laboratories, testing centers, inspection facilities and other support establishments in accordance with national and international standards.
3. Organizations and individuals recognized as members of AI linkage clusters are entitled to the following incentives:
a) Be prioritized to access and use national AI infrastructure, shared data and testing platforms at preferential costs;
b) Be supported to participate in programs for human resource training, trade promotion, and tasks related to science, technology and innovation in key fields.
4. The Government shall elaborate criteria, procedures for recognition, operational mechanisms of AI linkage clusters, and incentives prescribed in clause 3 of this Article.
Article 25. Supporting enterprises in AI
1. Startups, small and medium enterprises shall be supported with costs for conformity evaluation as prescribed in this Law; provided with free sample dossiers, self-evaluation tools, training and consultancy; and prioritized for support from the Fund.
2. Startups, small and medium enterprises, scientific and technological organizations, and research groups having feasible innovation projects shall apply for support tickets to use computing infrastructure, shared data, large language models in Vietnamese and ethnic minority languages, training platforms, testing environments, and technical consulting services for research, development and deployment of AI.
3. Enterprises with research, development and innovation capacity in AI shall be prioritized to participate in tasks under the national science, technology and innovation program; investment tasks for development of prioritized high-tech, strategic technologies, key digital technology products and services; and supported in developing core technologies, platform models, hardware and high-performance training technologies in accordance with the national AI capacity development orientation.
4. Enterprises participating in AI regulatory sandboxes shall be supported with technical consultancy, risk evaluation, safety testing and connection to testing and inspection facilities as prescribed by laws.
5. Enterprises sharing data, models, tools or research results for AI development are entitled to incentives or support as prescribed by laws in accordance with laws on data, protection of personal data and intellectual property.
6. The State shall encourage cooperation between enterprises, research institutes, universities and innovation centers in developing AI technologies, commercializing research results, and expanding innovation capacity; encourage enterprises to make long-term investment in researching and developing AI.
7. The Government shall elaborate mechanisms, policies, requirements and procedures for supporting enterprises in AI.
ETHICS AND RESPONSIBILITIES IN AI ACTIVITIES
Article 26. National AI ethics framework
1. The National AI ethics framework shall be issued in accordance with the following principles:
a) Ensure the safety, reliability; cause no harm to human life, health, honor, dignity and mental well-being;
b) Respect human rights and civil rights; ensure fairness, transparency and non-discrimination in development and use of AI;
c) Promote happiness, prosperity and sustainable development of individuals, communities and society;
d) Encourage innovation and social responsibility in research, development and application of AI.
2. The National AI ethics framework shall be reviewed and updated periodically, or when there is a significant change to technology, laws or management practices.
3. The National AI ethics framework serves as the basis for guiding the development of standards, technical regulations, professional guidelines and policies encouraging safe, reliable and responsible AI development.
4. The State shall encourage organizations and individuals to apply the National AI ethics framework during the research, development, provision, deployment and use of AI systems to ensure the transparency, equality, safety and respect for human rights.
5. The Minister of Science and Technology shall promulgate the National AI ethics framework in accordance with clause 1 of this Article.
Article 27. Ethical responsibilities and impact evaluation regarding application of AI in state management and public service provision
1. The use of AI systems in state management and public service provision shall ensure transparency, openness, and comply with the National AI ethics framework.
2. AI systems shall not replace the authority and decision-making responsibilities of decision-makers as prescribed by laws. Decision-makers are responsible for reviewing and using the results provided by AI systems.
3. Authorities operating high-risk AI systems or systems that significantly impact human rights, social equality or public interests shall prepare conformity evaluation reports on the use of such systems. The reports include risk identification, control measures, human oversight and intervention capabilities.
4. Authorities preparing conformity evaluation reports are responsible for the content, truthfulness and completeness of the reports. The reports shall be disclosed as prescribed by laws, except for information classified as state secrets, business secrets or personal data.
5. The Government shall elaborate the content, procedures and responsibilities for impact evaluation, risk management and supervision of the use of AI systems in state management and public services.
INSPECTION AND VIOLATION HANDLING
1. The inspection of AI activities shall comply with the laws on inspection.
2. Authorities and organizations taking charge of state management of AI shall conduct inspection on the compliance of organizations and individuals in AI activities.
3. During the inspection, relevant organizations and individuals shall provide technical dossiers, activity logs, training data and essential information for determination of causes of violations, incidents or determination of responsibilities. The provision of information shall comply with the laws on protection of state secrets, data, personal data and intellectual property.
4. The inspection conclusion and decision on imposition of administrative fines shall be disclosed as prescribed by laws.
Article 29. Handling of violations and paying damages
1. Organizations and individuals violating this Law and other laws related to AI shall be subject to administrative fines or criminal prosecution depending on the severity and consequences of the violations. They shall also pay damages in accordance with civil laws.
2. In cases where a high-risk AI system is managed, operated and used in accordance with the laws but still causes damage, the deployer shall pay damages for the affected person. After paying damages, the deployer may request reimbursement from the provider, developer or relevant parties if there is an agreement between the parties.
3. The responsibility for paying damages prescribed in clause 2 of this Article shall be exempt in the following cases:
a) The damage occurs due to the intentional fault of the damaged party;
b) The damage occurs due to force majeure or emergency circumstances, unless prescribed otherwise.
4. In cases where an AI system is illegally accessed, controlled or interfered with by a third party, the third party is liable for paying damages. If the deployer or provider is at fault for allowing the system to be accessed, controlled or interfered against the laws, they are jointly liable for paying damages in accordance with civil laws.
5. The State shall elaborate administrative fines for violations caused by AI systems.
Article 30. Content and responsibilities for state management of AI
1. The scope of state management of AI includes:
a) Developing, promulgating and organizing the implementation of strategies, policies, programs and legislative documents on AI;
b) Promulgating and organizing the implementation of standards and technical regulations on AI;
c) Managing, coordinating and developing the national AI infrastructure;
d) Managing and supervising AI activities;
dd) Disseminating policies, laws; statistics, reports, scientific research, international cooperation in AI;
e) Conducting inspections; handling violations; resolving disputes, complaints and denunciations related to AI.
2. Responsibilities for state management of AI:
a) The Government shall unify the state management of AI;
b) The Ministry of Science and Technology is the focal point accountable to the Government for nationwide state management of AI;
c) Ministries, ministerial authorities, within their functions, tasks and jurisdiction, shall cooperate with the Ministry of Science and Technology in performing state management of AI;
d) Province-level People’s Committees shall perform state management of AI in provinces.
Article 31. Principles for providing information and data for state management
1. Competent state authorities, organizations and individuals assigned to perform state management as prescribed in this Law are responsible for ensuring the confidentiality of information and data, as well as business secrets during the task implementation including technical dossiers, training data, source code and algorithms as prescribed by laws.
2. The requests for information and data from organizations and individuals shall only be made when necessary in accordance with the scope, purpose and content of state management activities.
3. Provided information and data must be protected in accordance with the laws.
Article 32. International cooperation
1. International cooperation in AI shall comply with laws on science and technology, laws on technology transfer, other relevant laws and international treaties of which the Socialist Republic of Vietnam is a member.
2. The State shall encourage the international cooperation in sharing high-performance computing infrastructure, data, human resources, scientific research and recognizing conformity evaluation results as prescribed in this Law.
Article 33. Annulment of some chapters, articles, clauses and points of the Law on Digital Technology Industry No. 71/2025/QH15
Clause 9, Article 3; clause 7, Article 4; clause 6, Article 12; point dd, clause 2, Article 34 and Chapter IV of the Law on Digital Technology Industry are annulled.
This Law comes into force from March 01, 2026, except for Article 35 of this Law.
Article 35. Transitional provision
1. For AI systems that have been put into operation before the effective date of this Law, providers and deployers are responsible for fulfilling obligations prescribed in this Law within the following time limits:
a) 18 months from the effective date of this Law for AI systems in the fields of health, education and finance;
b) 12 months from the effective date of this Law for AI systems not prescribed in point a of this clause.
2. Within the time limits prescribed in clause 1 of this Article, AI systems may continue to operate, except for cases where state management authorities for AI request suspension or termination of operations if they determine that such systems poses a risk of causing serious damage.
This Law was ratified by the 10th meeting of the 15th National Assembly of Socialist Republic of Vietnam on December 10, 2025.
|
PRESIDENT OF THE NATIONAL ASSEMBLY Tran Thanh Man |
